Skip to main content

Full text of "Transactions of the Royal Society of Edinburgh"

See other formats


H-.C 



a* 



TRANSACTIONS 



OF THE 



BOYAL SOCIETY OF EDINBURGH. 



SM*. G.Jffr. 



TRANSACTIONS 



OP THE 



ROYAL SOCIETY 



OF 



EDINBURGH. 



VOL. XLVII. 




EDINBURGH : 

PUBLISHED BY ROBERT GRANT & SON, 107 PRINCES STREET, 
AND WILLIAMS & NORGATE, 14 HENRIETTA STREET, COVENT GARDEN, LONDON. 



MDCCCCXI. 



No. 



I. 


Published 


April 15, 1909. 


No. 


XV. 


II. 


>> 


May 3, 1909. 


>> 


XVI. 


III. 


>) 


May 25, 1909. 


)> 


XVII. 


IV. 


>) 


May 27, 1909. 


>) 


XVIII. 


V. 


!) 


June 9, 1909. 


>) 


XIX. 


VI. 


,) 


September 13, 1909. 


jj 


XX. 


VII. 


)) 


September 24, 1 909. 


m 


XXI. 


VIII. 


)) 


October 12, 1909. 


)> 


XXII. 


IX. 


)> 


October 13, 1909. 


» 


XXI II. 


X. 


>1 


October 21, 1909. 


)! 


XXIV. 


XI. 


>) 


December 31, 1909. 


» 


XXV. 


XII. 


;> 


January 7, 1910. 


>5 


XXVI. 


XIII. 


!' 


January 29, 1910. 


J) 


XXVII. 


XIV. 


)j 


February 1, 1910. 


!> 


XXVIII 



Published 



February 8, 1910. 
March 3, 1910. 
March 24, 1910. 
August 30, 1910. 
September 21, 1910. 
January 26, 1911. 
January 13, 1911. 
January 21, 1911. 
February 10, 1911. 
February 4, 1911. 
March 17, 1911. 
March 31, 1911. 
April 18, 1911. 
April 18, 1911. 



CONTENTS. 



PART I. (1908-09.) 

NUMBER PAGE 

I. Notes on Hydrodynamics, chiefly on Vortex Motion. By Professor 

Andrew Gray, F.R.S., . . . . . .1 

II. The Glacial Deposits of Western Carnarvonshire. By T. J. Jehu, 
M.D. (Edin.), M.A. (Camb.), F.G.S., Lecturer in Geology at the 
University of St Andrews. (With Four Plates and One Map), . 17 

III. Scottish National Antarctic Expedition : Observations on the 

Anatomy of the Weddell Seal (Leptonychotes Weddelli). By David 
Hepburn, M.D., Professor of Anatomy, University College, Cardiff 
(University of Wales). (With One Plate), . . .57 

IV. Supplementary Report on the Hydroids of the Scottish National 

Antarctic Expedition. By James Ritchie, M.A., B.Sc, Natural 
History Department, The Royal Scottish Museum, . . .65 

V. The Meteorology of the Weddell Quadrant and Adjacent Areas. By 

Robert C. Mossman. (With Five Plates), . . . .103 

VI. TJte Monsoons of the Chilian Littoral (Preliminary Note). By 

R. C. Mossman, . . . . . . .137 

VII. A Carboniferous Fauna from Noivaja Semlja, collected by Dr W. S. 
Bruce. By G. W. Lee, D.Sc, H.M. Geological Survey. With 
Notes on the Corals by R. G. Carruthers. (With Two Plates), . 143 

VIII. Scottish National Antarctic Expedition: Osteology of Antarctic Seals. 
By Robert B. Thomson, M.B., Ch.B., University of Edinburgh. 
(With One Plate), , . , , . .187 



\/ 



vi CONTENTS. 



PART II. (1909-10.) 

1UMBKR PAGE 

IX. Lower Palceozoic Hyolithidce from Girvan. By F. R. Cowper Reed, 

M.A., F.G.S. (With Three Plates), . . . .203 

X. Report upon the Anatomy and Embryology of the Penguins collected 
by the Scottish National Antarctic Expedition, comprising : (1) Some 
Features in the Anatomy of the Penguin; (2) The Embryology of 
the Penguin : a Study in Embryonic Regression and Progression. 
By D. Waterston, M.A., M.D., F.R.C.S.E., and A. Campbell Geddes, 
M.D. (With Three Plates), . . . . . ,223 

XL The Pharmacological Action of Harmaline. By James A. Gunn, 

M.A., M.D., D.Sc, . ..... 245 

XII. The Pharmacological Action of Protocatechyl-Tropeine. By Professor 

C. R. Marshall, ....... 273 

XIII. The Pharmacological Action of Tutu, the Toot Plant of New Zealand. 

By Professor C. R. Marshall, . .287 

XIV. On the Life-history of Hydrobius fuscipes, L. By Frank Balfour- 

Brownk, M.A. (Oxon.), F.Z.S. (With Three Plates), . . 317 

XV. Strophanthus sarmentosus : its Pharmacological Action and its Use as 
an Arrow-Poison. By Sir Thomas R. Fraser, M.D., F.R.SS. L. & E., 
Professor of Materia Medica in the University of Edinburgh, 
and Alister T. Mackenzie, M.A., M.B., Ch.B., Carnegie Research 
Scholar. (With Eleven Plates), . . . . .341 



PART III. (1910-11.) 

XVI. The Aborigines of Tasmania. Part II. — The Skeleton. By Principal 
Sir William Turner, K.C.B., F.R.S., President of the Society. 
(With Two Plates), . . . . . .411 

XVI I. On the Fossil Osmundacece. Part IV. By R. Kidston, LL.D., F.R.S., 
F.G.S., Foreign Mem. K. Mineral. Gesell. zu St Petersburg, Hon. 
Sec. R.S.E. ; and D. T. Gwynne-Vaughan, M.A., F.L.S., Professor 
of Botany, Queen's University, Belfast. (Plates I. -IV.), . . 455 



CONTENTS. vii 

NUMBER PAQK 

XVIII. The Lamellibranchs of the Silurian Rocks of Girvan. By Wheelton 

Hind, M.D., B.S., F.R.C.S., F.G.S. (With Five Plates), . . 479 

XIX. TJie Alcyonaria of the Cape of Good Hope and Natal. Alcyonacea. 

By J. Stuart Thomson, Ph.D., F.L.S. (With Four Plates), . 549 

XX. TJie Phase of , the Nucleus known as Synapsis. By A. Anstruther 
Lawson, Ph.D., D.Sc, F.L.S. , F.R.S.E., Lecturer in Botany, 
University of Glasgow. (With Two Plates), . . . .591 

XXL Observations on the Body Temperature of the Domestic Fowl ( Gallus 

gallus) during Incubation. By Sutherland Simpson, M.D., . 605 



PART IV. (1910-11.) 

XXII. The Temperature Seiche. Part I. — Temperature Observations in the 
Madiisee, Pomerania. Part II. — Hydrodynamical Theory of 
Temperature Oscillations in Lakes. Part III. — -Calculation of the 
Period of the Temperature Seiche in the Madiisee. By E. M. 
Wedderburn, W.S. Part IV. — Experimental Verification of the 
Hydrodynamical Theory of Temperature Seiches. By E. M. 
Wedderburn, W.S., and A. M. Williams, M.A., B.Sc. (With 
One Plate), . . . . . . . .619 

XXIII. TJie Jurassic Flora of Sutherland. By A. C. Seward, F.R.S., 

Professor of Botany, Cambridge. (Plates I.-X. ; text-figures 1-14 ; 
sketch-maps, and views of the coast), .... 643 

XXIV. On the Structure and Affinities of Diplolabis romeri (Solms). By 

W. T. Gordon, M.A., B.Sc, B.A., Falconer Fellow of Edinburgh 
University, Lecturer in Palaeontology, Edinburgh University. 
(With Four Plates), . . . . . .711 

XXV. Thermo- Electric Diagram from —200° C. to 100 c C, deduced from the 
observations of Professors Dewar and Fleming. By J. D. Hamilton 
Dickson, M.A., Peterhouse, Cambridge, .... 737 

XXVI. The Plant Remains in the Scottish Peat Mosses. Part IV. By 
Francis J. Lewis, M.Sc, F.L.S., Lecturer in Geographical Botany, 
University of Liverpool. (With Five Plates), . . . 793 

(^jWy/Vl A S WixJ^Vv (f) M^T Hn&) WW °$ r Z*JL** I ) 



vin CONTENTS. 

NtMHKl; PAGK 

XXVII. On an Entoproctan Polyzoon [Barentsia benedeni) new to the British 
Fauna, with Remarks on Related Species. By James Ritchie, 
M.A., B.Sc, Natural History Department, The Royal Scottish 
Museum. (With One Plate), . . . . .835 

XXVIII. An Investigation into the Effects of Errors in Surveying. By Henry 

Briggs, B.Sc., A.R.S.M., . . .849 

Index, .......... 879 




TRANSACTIONS 



OF THE 



ROYAL SOCIETY OF EDINBURGH. 

VOLUME XLVII. PART L— SESSION 1908-9. 




CONTENTS. 

I. Notes on Hydrodynamics, chiefly on Vortex Motion. By Professor Andrew Gray, F.R.S., . 1 

(Issued April 15, 1909.) 

II. The Glacial Deposits of Western Carnarvonshire. By T. J. Jehu, M.D. (Edin.), M.A. (Camb.), 
F.G.S., Lecturer in Geology at the University of St Andrews. (With Four Plates and 
One Map), .......... 17 

(Issued May 3, 1909.) 

X HI. Scottish National Antarctic Expedition: Observations on the Anatomy of the Weddell Seal 
(Leptonychotes Weddelli). By David Hepburn, M.D., Professor of Anatomy, University 
College, Cardiff (University of Wales). (With One Plate), . . . .57 

(Issued May 25, 1909.) 

IV. Supplementary Report on the Hydroids of the Scottish National Antarctic Expedition. By 

James Ritchie, M.A., B.Sc, Natural History Department, The Royal Scottish Museum, 65 

(Issued May 27, 1909.) 

V. The Meteorology of the Weddell Quadrant and Adjacent Areas. By Robert C. Mobsman. 

(With Five Plates), ......... 103 

(Issued June 9, 1909.) 

VI. The Monsoons of the Chilian Littoral (Preliminary Note). By R. C. Mossman, . .137 

(Issued September 13, 1909.) 

VII. A Carboniferous Fauna from Nowaja Semlja, collected by Dr W. S. Bruce. By G. W. Lee, 
D.Sc, H.M. Geological Survey. With Notes on the Corals by R. G. Carruthers. 
(With Two Plates), . . . . . . . . .143 

(Issued September 24, 1909.) 

X, VIII. Scottish National Antarctic Expedition : Osteology of Antarctic Seals. By Robert B. 

Thomson, M.B., Ch.B., University of Edinburgh. (With One Plate), . . .187 

(Issued October 12, 1909.) 



EDINBURGH: 

PUBLISHED BY ROBERT GRANT & SON, 107 PRINCES STREET, 

AND WILLIAMS & NORGATE, 14 HENRIETTA STREET, COVENT GARDEN, LONDOJN. 



MDGCCCIX. 

Price Nineteen Shillings and Ninepence. 



TRANSACTIONS. 



I. — Notes on Hydrodynamics, chiefly on Vortex Motion. 
By Professor Andrew Gray, F.E.S. 

(MS. received November 6, 1908. Read November 16, 1908. Issued separately April 15, 1909.) 

1. In the teaching of hydrodynamics many points of difficulty arise, both for 
teacher and for student. The subject abounds in subtleties even in its very elements, 
and the advanced student frequently finds himself in a state of doubt as to 
fundamental questions which crop up unexpectedly in connection with various 
problems. The following notes contain a discussion of a few of these fundamental 
matters : for example, they deal with some theorems of energy which have been 
found difficult by students, perhaps mainly through want of perfectly explicit state- 
ment of their scope and purpose. Finally will be demonstrated a theorem of 
vortex-motion, particular cases of which have been given by various writers, but 
which I have not seen stated elsewhere in the same generality. This will be found 
to lead to Lord Kelvin's well-known and far-reaching theorem of circulation, and 
to other theorems of the vortex-motion of a perfect fluid, some of which are already 
known. 

2. First it may be recalled that if q be the resultant velocity of a particle of 
the fluid at any point P at time t (or indeed any other quantity characteristic of 
an element of the fluid in motion), and ds the element of path actually described 
by the particle in the interval of time dt beginning at t, the acceleration of the 
particle is 

dt A ds 

This, of course, is the usual expression from which the component accelerations 
parallel to the axes are deduced. 

But if ds' be an element of a line drawn from P at any angle to ds, and q' the 
component of velocity in the direction of ds', the acceleration in this direction is 

dt *ds 
This is a very useful expression, and will be of great service in what follows. By 
putting u, v, w for q' we obtain easily the usual component accelerations. That the 

acceleration along ds' has this value is at once evident from the fact that, at time t, 
y TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. I). 1 



2 PROFESSOR ANDREW GRAY, 

the velocity in the direction of ds' at a point Q, at a distance ds in advance of P. 
along the line of flow, is q' + dq'/ds . ds. After the lapse of the interval dt, the 
particle which was at P has reached Q, and its velocity (along ds') has become, 
to the first order of small quantities, q' + dq'/dt . dt + dq'jds . ds. Since q = ds/dt, 
the particle when at P is thus gaining velocity in the direction of ds' at rate 

dq dq dq' 

0* = ¥ + «W .... (i). 

3. An elementary point of some difficulty to the student arises in this connection. 
The integral 

J dt 
is to be differentiated with respect to ds', specified as in § 2. The result is 

*&*-%*-% .... (2). 

ds J dt dt ds dt 

Here we have ds/ds' = cos 0, where 6 is the angle between ds' and the terminal element 
of the line of integration. It is at first a little difficult to see that ds/ds' has this 
value. But the whole question is, What is the change in fdq/dt . ds produced by 
a small step ds' inclined at an angle 6 to ds ? Now the step ds' may be regarded 
as made up of a step ds' cos along the stream-line, followed by one ds' sin 6 
at right angles to the stream -line. The former gives the element dq /dt . ds' cos 6 of 
the integral ; the latter, being at right angles to the stream-line, leaves the integral 
unchanged ; therefore the result of the differentiation is dq/dt . cos 0, or dq'/dt. 

The student is at first tempted to take ds' as one component of a step ds along 
the stream-line, that is, as ds cos 6. He forgets that if this is done the element 
dq/dt . ds of the integral, which he now makes to correspond to ds', is the result, not 
of the step ds' alone, but in part also of the coexistent step ds sin 6. 

4. The flow of energy in the hydrodynamic field may be discussed in two ways. 
We may determine the rate of change of energy within a closed surface S fixed 
in the fluid, or we may find the rate of change of the energy of the definite mass 
of fluid which at time t occupies the space within the given surface S. These are 
two distinct problems — since at time t + dt the fluid which at time t occupied the 
space within the surface no longer does so precisely — and a comparison of their 
solutions is instructive. As we shall see presently, the solution of one problem 
can be derived from that of the other ; and the application of the results to 
particular cases gives various well-known hydrodynamical theorems. 

5. We take first the problem of the rate of change of the energy of the mass 
of fluid contained at time t within the surface S. Let q denote, as usual, the 
resultant velocity of a particle of the fluid, and q its total time-rate of variation. 
If, as is usually the case, the external forces (per unit mass) be derivable from a 
potential — which we shall call V — we have 

3V q dp 
qq = -q — - J- £-. 
HH 'ds p ds 



NOTES ON HYDRODYNAMICS, CHIEFLY ON VORTEX-MOTION. 3 

Multiplying both sides by dm, an element of the fluid mass, and integrating for 
the whole mass of fluid within the surface, we obtain 

jqqdm = - jq—dm- \q^-d~& . . . (3), 

if dzj be the volume (dm/p) of the element of mass dm. Now usually V is 
independent of t, and therefore dV/dt = 0. Thus for qdY/ds in the first integrand 
on the right we can write dV/dt + qdY/ds, which is clearly dV/dt. Thus if T 
denote the kinetic energy of the fluid which fills the surface at time t, and E 
denote the potential energy from which the external forces are derived for the 

same fluid mass, we have 

f ., dT fdV, dE 
qqdm — — , \ — dm = — . 
J dt' J dt dt 

The equation found above can therefore be written in the form 

*(T + 1>~/4E*, .... (4). 

This gives, in an exceedingly compact form, the value of the total time-rate of 
variation of the sum of the kinetic and potential energies T and B. 

6. The integral on the right can be transformed by partial integration, without 
the introduction of Cartesian co-ordinates, by proceeding as follows. The continuity 
of the motion from point to point in the fluid involves the possibility of dividing 
the fluid mass up into narrow tubular portions bounded by non-intersecting stream- 
lines. We shall call these tubes of flow. Their distribution is determinate for any 
instant (or point of time), but varies in the general case from instant to instant. 
Each such tube (unless it be endless, and entirely contained within the surface) will 
enter the closed surface S at one element c/S x of surface, and emerge at another 
element cZS 2 . We shall denote for any element of one of these tubes, say an element 
of length ds, the average area of cross-section by a- ; then <r varies along the tube, 
that is, it is to be regarded as a definite function of s, at each instant of time. 
The element of volume dzs will now be <rds. 

7. First, then, integrating the expression on the right of (4) by parts, we get as the 
integrated terms for a single tube —p<>CL2' J 2~' r P\ ( li a '\-> wnere °"i> °"2 are ^ ne cross-sections 
of the tube at entrance into and emergence from the surface S. If q n he the com- 
ponent of velocity normal to the surface inward from an element c/S of the surface, 
this sum of terms may be written (pq n d$) 2 + (pq n d$) v The aggregate result for all 
the tubes crossing the surface is the integral 

Jpq n dS 

taken over the surface S. Thus we obtain, since dw = <rds, 

^(T + E ) = jpq n dS + jp l - d -{q<r)dZZ 



4 PROFESSOR ANDREW GRAY, 

If there be endless tubes of flow contained entirely within the surface, no change 
in the surface integral in this result is required. The volume integral on the right, 
however, must be evaluated for each such tube. 

The expression 

dq q da- 
ds o- 3s 

is the "divergence" of the fluid within a tube of flow, that is, the time-rate of 
increase of the volume of the fluid, per unit of volume, at the element ds. This 
expression is equivalent to 

du dv dw 

dx dy r)z' 



the divergence expressed in terms of the component velocities with reference to 
rectangular axes. [It is not difficult to verify this equivalence.] 
Equation (5) may thus be written in the form 



(T + E) = fp(lu + mv + nw)dS + j f fpP~ + ^ + d ^\dxdydz . (6), 



where I, m, n are the direction-cosines of the inward drawn normal to the element 
of surface rfS. 

8. From the equation (5) [or (6)] we can deduce the value of 



f(yq 2 + pV)dn, 



that is, of the time-rate of variation of the energy within the fixed surface S. 

Let U denote the volume of the space within S, then this at time t is the volume 

of the mass of fluid which we have been considering. Let also e denote the average 

value of the sum of the kinetic and potential energies per unit of volume ; then for the 

whole mass of the fluid 

T + E = Ue. 
Thus 



therefore 



-(T + E) = e— + U-; 
dV ' dt dt' 



4r> +E >-4; 



for U, considered as the volume of a definite mass of fluid, is subject to change. This 
by § 7 is the same thing as the equation 

-f(^/ + ,V)(| + ?|), to . (7). 



But by >j 2 



|< W + P V) = 1(| W 2 + P Y) + q^pq* + P Y). 



NOTES ON HYDRODYNAMICS, CHIEFLY ON VORTEX-MOTION. 5 

Thus we obtain 

Remembering that {dq/ds + (ql<r)d<r/ds}dvJ = d(q<r)/ds . ds, and integrating the second line 
on the right by parts as in § 7, we obtain 

j?(hpT + pV)dK = J \$tf + pV +p)q n dS + fp(% + i*£)dt3 . . (8). 

The integral on the left of (8) is the rate at which the sum of the kinetic and 
potential energies within the surface is increasing ; and the integrals on the right show 
that this rate is equal to the rate of flow of energy into the space across the surface, 
together with the rate at which work is done by pressure in consequence of expansion 
of the fluid within the space and of passage of fluid across the surface. 

9. The value of the divergence found above for a tube of flow leads to a corre- 
sponding mode of expressing the equation of continuity of the fluid. If p be the 
density of the fluid at the element of volume dm, then, since the increase of volume in 
time dt is (dq/ds + q/a- .d<r/ds)dt per unit volume, the mass for this at density p is 
p(dq/ds + q/cr . da-jds)dt. But since the mass of the element is not changed, for it 
preserves its identity, this expression must have the value — dp/dt . dt. Thus we 
obtain the equation of continuity in the form 



dt H \ds a- ds) 



or 

dp dp 



(9). 



Op dp /da q d<r\ 

If we put for the divergence, this equation can be written 

f + P© = (10). 

at 

10. As an example of these results we apply them to the steady motion of the fluid 
within a portion of a tube of flow bounded by cross-sections perpendicular to the 
stream-lines. In this case the surface-integral on the right is limited to the ends of 
the tube. Let dS l5 c/S 2 be the areas of the ends, and q v q 2 be the velocities there 
along the tube. Then the surface integral becomes 

(hi + Vi + P A PifcdSx - (te\ + v 2 +^Ws 2 . 
\ pj \ Pi/ 

But since the motion is steady, p^d^ = p 2 q 2 d$ 2 , since the former is the mass of fluid 



6 PROFESSOR ANDREW ORAY, 

which Hows into the tube, and the latter the mass which flows out, per unit of time. 
Denoting this constant quantity by m, we obtain 

kz+pV+p^s^fe + v.+ft-^-v,-^ . . (ii). 
j \ p x p 2 / 

Now consider the volume-integral on the right of (8). By the equation of con- 
tinuity given in § 9 we have 

{p®dvs=-Jp d .P . . . (12), 

J j p~ 

where the integral on the right is to be taken along the tube here considered, in the 
direction of flow. Integrating by parts, we find 

J \ds <t 3s/ \jOjj p 1 J Pi p J 

This added to the former result gives 

fiW+P^+PhndS + fp^ + ^d^^^yi + Y.-hl-Y,- f"^) . (13). 

Thus, since the left-hand side of (13) vanishes in the case of steady motion, we 
obtain finally for any cross-section of the tube the well-known equation 

k 2 + V+ f^ = const. . . . . (14), 

J P 

which is generally obtained by another process. It is to be understood that the 
integral, fdp/p, is to be taken along the stream-line, from any chosen starting-point 
up to the cross-section to which q and V belong. It is thus shown that ^ 2 + V +fdpjp 
is constant along a stream-line ; but its constant value, it is to be observed, may change 
from one stream-line to another. 

11. We shall now find an expression for 



/: 



&* + tf + T + |* 



in the general case. This will show that when the fluid is destitute of what may be 
called elemental rotation the value of this expression, which is a function of the time, 
is the same throughout the fluid at any one instant, and becomes an absolute constant 
when the motion is steady. 

The theorem thus arrived at will be found to lead at once to Lord Kelvin's 
theorem of circulation, from which the permanence of vorticity, or non-vorticity, in any 
portion of a perfect fluid, can be inferred. It will be shown also how another 
relation can be obtained, which affords another view of the proof of the theorem of 
permanence. 

Let \j/- denote 



NOTES ON HYDRODYNAMICS, CHIEFLY ON VORTEX-MOTION. 



where ds is in the direction of the resultant velocity q, and p is as usual a function 
of p. The integrals are supposed taken along the stream-line from any chosen 
starting-point to the point considered. If now we differentiate in any direction 
inclined at an angle to the direction of ds and q, we get 



But we know that 



ds dt ds' ds' p ds' 



d l +q d i + d I+l d p=o 

dt H ds ds' P ds' 



for the first two terms make up dq'/dt, 
and this, by the equation of motion for 
the direction ds', is — (dV/ds' + 1 jp . dp/ds'). 
Thus by subtraction we get 



OS \os OS / 



But if u> gg , denote the component of ele- 
mental rotation of the fluid about the 
normal to the plane of ds and ds', at P 
their point of meeting, as indicated in the 
figure, we can easily prove that 




- 2o)„, sin 0- 



dq 

: 8? 



d£ 

ds 



so that 



d\j/ 



= - 2w ss , q sin i 



(15), 



(16). 



12. I have not seen this theorem stated before, though various particular cases 
of it are known from which it might be inferred. It asserts that, at time t, \J/- is 
constant along a stream-line, and likewise along any line ds' so drawn that « M . is 
zero for the plane determined by ds' and ds. \|/- is in general a function of t, and 
the theorem shows how its variation from point to point depends on the motion of 
the fluid. 

By turning ds', without altering 6, we can change the plane of ds and ds' from that 
for which <a gl , sin 6 is zero to that — inclined to the former at an angle of 90° — for which 
w gg < sin 6 is a maximum. The normal to the latter plane may or may not be the 
direction of the axis of resultant rotation at P (see figure). 

13. Equation (15) may be regarded as an equation of motion for the direction ds' 
inclined at the angle 6 to ds, the direction of flow. If we take ds' parallel to the fixed 
axes Ox, Oy, Oz in succession, we obtain three equations : — 



^ + 2a> sx qsm0 sx = O, 



~- 1- 2w sy q sin sy = 



dxh 

■^ + 2o} sZ q sin 6 SZ = 



(17). 



8 PROFESSOR ANDREW GRAY, 

If, according to the usual notation, we put 



these equations become 



dw dv _du dw os-_3'"' du 

dp dz' dz dx dx dy 



°£ + 2w v -2vZ = 0, ^ + 2^-2^ = 



• (18), 



which may be regarded as the Cartesian equations of motion of the fluid. These 
equations are given in Lamb's Hydrodynamics, 3rd edition, § 146, but are derived 
directly from the ordinary Eulerian equations of motion. 

14. The proof of equation (15) deserves a little attention. We can show that the 
equation holds, and also make it clear that w ss , is an angular velocity of rotation — 
a point that is not always brought out in the kinematics of vortex-motion — by pro- 
ceeding as follows. The average velocities along the four sides of the small parallelo- 
gram of the figure may be taken to be the velocities of the fluid in these directions at 
the middle points of the sides. These are, for the sides PQ, QR, RQ', Q'P, 

1 ^1 7 I dl l 7 M T 

9 + hfods, 2+ & *> + *&*»> 



/ ,d</ 3<7 A f . , W , A 



Multiplying the first and third of these by ds, and the second and fourth by ds', and 
adding, we obtain 



3s 3s'/ 

Thus if q c be the component of velocity along the boundary of the parallelogram 
at any element dc, we have for the parallelogram 

I qjlc = (JL — -2, )dsds. 
J ic \ds dsj 

But the angular velocity of a fluid particle at dc about a point, P" say, in the 
plane of ds and ds' and within the parallelogram, is qjp, where p is the length of 
the perpendicular from P" on the side upon which dc is situated. Thus 



{'h p dc=( d l-^\d,ls. 
J j? \3s ds'J 



If co„, be the mean angular velocity for particles at different points of the periphery 
of the parallelogram, the integral on the left is evidently 2w sg -dsds' sin 6. Thus 
we obtain, as stated in (15), 



NOTES ON HYDRODYNAMICS, CHIEFLY ON VORTEX- MOTION. 



15. Returning now to equation (16), we write instead of \f/, x + \dq/dt . ds, 
so that 



J p 



Equation (16) may thus be written 



-A = - 2a)„, a sin - -±- 

ds s l dt 



(19). 



(20). 



As in the diagram, let AB be any line drawn in the 
fluid, and PQ be the element ds', which we suppose to be 
on this line. 

As before, q is the resultant velocity of the fluid at 
P, and q' = q cos 6 is the component at P along the tangent 
to AB at P. Now q sin 6 is the velocity at P at right 
angles to ds', that is, q sin is the velocity with which 
each point of the path AB is moving at P towards the 
right in the plane of the paper. The product q sin 6ds' 
is therefore, the rate at which any circuit including ds' 

is increasing in area, in consequence of the motion of ds' at right angles to itself, 
in the plane of the paper, without alteration of length. 

Multiplying (20) by ds', and integrating from A to B, we obtain 




x-x = 



f f W 

I 2o) ss ,q sin 6 ds - I ~rd$ 



(21). 



We see that the first integral on the right is twice the rate at which the surface 
integral of elemental rotation, taken over any surface of which AB is part of the 
boundary, is changing in consequence of the fact that each element ds' is being 
carried at right angles to itself by the motion of the fluid. It is also, it is to be 
noted, the rate at which the flow of the fluid along AB is changing from this cause. 

16. But the area enclosed is also changing in consequence of the motion of one end 
of the element ds' relatively to the other. The surface integral is changing from this 
cause also, and the rate of change can be found most easily by calculating the rate of 
increase of flow along AB arising from the change of length of the element. 

To do this, consider the new position of ds' after the lapse of time dt. Let the 
plane of the initial position of ds', and of the displacement qdt of P, be taken as 
a plane of reference. The other end Q of ds' has been displaced through a distance 
(q + dq/ds' . ds')dt which is not in this plane, and is not parallel to the displacement 
qdt of P. The increase of the rate of flow of fluid along ds' arising from the 
difference (dq/ds' . ds')dt is qdq/ds' . ds'. 

If this is not intuitively perceived, it can be proved as follows : — 

The component of the displacement (q + dq/ds'. ds')dt of Q along ds' is 
(q + dq/ds' . ds')dt . cos (6 + dd), or (q cos 6-qdd. sin + dq/ds' . ds' cos d)dt. Thus the end 
TRANS. ROY. SOC. EDIN., VOL XLVII. PART I. (NO. 1). 2 



10 PROFKSSOR ANDREW GRAY, 

Q suffers in dt a displacement relatively to P of amount (dq/ds' . ds' cos 6 — qd6 . sin 6)dt, 
parallel to its original length. 

The additional flow due to this displacement is (q' dq/ds' . ds cos 6 — qq'dd . sin 6)dt, or 



(q^ds' cos 2 - q 2 dO . sin cos \dt, 



since q = q cos ». 

But Q has also moved out through the distance (q + dq/ds' . ds')dt . sin (6 + d0) 
at right angles to the original position of ds'. This is not parallel to q' sin Odt, the 
displacement of P, but is so to the first order of small quantities. (This degree 
of approximation is, however, sufficient for the present calculation, as the velocity 
at Q in the direction at right angles to the plane ds, ds' is a small quantity of 
the first order. Thus the flow due to the deviation from the plane ds, ds' must 
be a small quantity of the second order.) We have to the degree of approximation 
required for the present purpose {(q + dq/ds' . ds') sin (6 + <W) — q sin 9\dt, or 

( q cos dO + --,ds' sin Ojdt 

for the only other relative displacement that need be considered. 

The velocity in the direction of this displacement is approximately q sin 6, and 
therefore the component of flow along the element due to the displacement is 

( q* sin 6 cos 6d0 + q&ds' sin 2 0\ dt. 

Thus the whole change in the rate of flow brought about by the two displacements is 

<?-i a!s'(sin 2 + cos 2 6) = q^,ds'. 

But the element ds', carried along by the motion of the fluid, is displaced as a whole 
in time dt through a certain distance parallel to its own direction, and this, apart from 
alteration of length or direction of the element, causes an alteration of the rate of flow 
along it. The amount of this is easily shown to be qdq/ds'.ds'. For the total growth 
in time dt of flow along ds', with its given length and direction, is (dq'/dt + qdq'/ds)ds'dt. 
This is due to the longitudinal and lateral displacements in dt taken together. But 
by (15), dq' /ds = dq/ds' + 2ft> gs . sin 6. Thus for the element ds', with its given length and 
direction, the rate of increase of flow is (dq'/dt + qdq/ds' + 2oo s /j sin 6)ds'. 

17. We therefore have finally 

(% - 2 q d l)ds' = - 2r 7w „ sin Ode' - 2<hds - %U . . (22). 

yds as / ds dt 

The expression on the right is twice the total time-rate of change of the surface- 
integral of rotation caused by the motion of ds', together with twice the time-rate 
of change of the same integral due to variation of velocity while ds' remains at rest. 
Integrating along AB we get 



* A -<flJ-flJ>- -2Jqo^ S m6ds'-2fq^ds'-f^ds' . (23). 



NOTES ON HYDRODYNAMICS, CHIEFLY ON VORTEX-MOTION. 11 

The first two terms on the right give twice the rate of change of the surface 
integral of elemental rotation (for any closed path of which AB forms part) due to 
the motion and the lengthening of the elements of AB ; the last gives twice the 
rate of change of this integral when the circuit is without motion. 

18. If we integrate round the closed path, then under the conditions already 
stated as the dependence of p upon p, and the existence of the potential V, we get 
from the vanishing of the left-hand side 

J2q^ sin eds +2 fq%h'+ f%ds =0 . . . (24), 

.' J os J dt 

and we see that the surface-integral of elemental rotation remains constant as the 
closed path moves with the fluid. 

It is to be noticed that for any closed path moving with the fluid 

finds' = - / 2^ sin 6d*' .... (24'). 
Equation (23) with the value of x inserted is 

Vb + / 7 - K - ( Va + / 7 - K) = ~ 2 / qojssS[n eds ' - 2 j qd S ds ' - !% ds ' ■ (25) ' 

B A AB 

and is equivalent to Lord Kelvin's theorem. The process of investigation here em- 
ployed, though longer than that usually adopted, has the advantage of tracing the 
various causes of change of the line-integral of flow along the path AB, and of throw- 
ing some additional light on the meaning of the theorem. 

19. If we take AB parallel to the axis of x, we obtain 

J 2q<o sll > sin 6ds' = - f (2v£ - 2 M r))dx . . . (26). 

AB AB 

Let us now consider a fluid in motion in virtue of vortices contained within it, and 
suppose that a surface S can be described within the fluid so that it is everywhere at 
a finite distance from the vortices. We can describe a closed path, consisting of two 
parts : (1) a part parallel to the axis of x, starting from a point A near the surface S, 
and ending at a point B, also near the surface, but remote from A ; and (2) a part 
starting from B and returning to A, but kept so near the surface everywhere as to 
avoid the vortices. If this path move with the fluid, then by (24/) 

/ 4t<1x+ I 2qo> sx sin 6 sx dx + \ ~ds' + J 2qu>$ s > sin Ods = . . (27). 

AB AB BA BA 

This equation can be written 

J d "dx - f (2v£ - 2wT))dx + f d -^dx + f 2qo> sli > sin 6ds' = . . (27'). 

AB AB BA BA 

But going back to (21), we see that the last two terms on the left are X B — X A , and 
as at every point of this part of the closed path &v = 0, we have 



X -X 

B A 



=/!*' • w 



BA 



12 



PROFESSOR ANDREW GRAY, 



If, further, the surface be so far from the vortices that the fluid is at rest [or, wh 
this is not the case, the motion be steady], we know that 

/¥*=»• 

Hence in either of these cases X A = X B . We have then 

I di dx ~ I ( 2r £~ 2wr >) d * =o 



en 



and we see that 



AB 



AB 



I jf^ dxd y dz - j f f( 2v £ - lio-qjdxdydz = 



(29); 



(30), 



where the volume integrals are taken throughout the fluid. In the case of steady 
motion this ffives 



Similarly, 



I f f (2v£ - 2wy } )dxdydz = 0" 
j f f(2w£ - 2uQdxdydz = 
f f f(2u v -2v£)dxdydz = 



(31). 



These equations are usually obtained only for the case of the fluid contained 
within a surface which is at every point infinitely distant from the vortex system. 
D Unless there is some error in the reasoning here stated this 
restriction would appear to be unnecessary. 

20. A vortex-line is defined as a curve the tangent to which 
at any point is in the direction of the resultant rotation at 
the point. Since d-^/ds' is zero when ds' is taken along a stream- 
line or along a vortex-line, it is zero for every direction in the 
plane determined at any point by the stream-line and the 
vortex-line through the point. It has therefore a maximum 
value at the point along a normal to the plane contained by 
these directions. 

A tubular space bounded by vortex-lines and filled with 
fluid in elemental rotation is called a vortex-tube. Consider 
a portion of a vortex-tube between two cross- sections formed by 
ds, an element of a stream- line, and ds f , taken at right angles 
to the plane of the stream-line and a vortex-line at a point 
on the surface of the tube. Such a portion of a tube is shown in the diagram by 
P'A' CD' BQ. P'A' and QB are taken at right angles to the planes of the stream- 
line and vortex-line at P' and Q, and represent ds' taken at these points. P'Q, 
A'B, CD' are vortex-lines, and P'C, QD' represent ds taken at the same points. 
Since ds is not necessarily at right angles to the vortex-line at P' or Q, let PC, QD 
be so taken. Then in PAC DBQ we have a portion of a vortex-tube with its 




NOTES ON HYDRODYNAMICS, CHIEFLY ON VORTEX-MOTION. 13 

ends perpendicular to vortex-lines, and PA, QB are the directions of els' at 
P and Q. 

Now for the plane P'A'C we have, since P'A' and P'C are at right angles, 

%= - 2w ss <g, 

3s 

or 

-^dt = - 2w 8 sds. 
ds 

Thus, for the surface integral of rotation over P'A'C we have 

J£dtds'=-±mrfdsds .... (32). 

ds 

But if w be the resultant angular velocity of rotation of the fluid about the vortex- 
line at P', and (p the angle between P'C (ds) and P'Q (the vortex-line), 

oj sin <f> = o)ss'> 
so that 

^^dtds' = - ±(i>ds ds. sin <f> . . . (33). 

OS 

But ^dsds'sin(f) is the area of the eud PAC of the portion of a vortex-tube. Calling 
this o-, we have 

gtMdt=-<*T .... (34). 

OS 

We can prove that this is the same for both sections of the tube. For, consider the 
circuit PABQP, made up of the vortex-lines PQ, AB, and the two intercepted lines ds', 
namely PA, QB. The line-integral 

reduces to d\js/ds'. ds' for PA, and a similar expression, with the minus sign prefixed, for 
BQ. But for any closed circuit moving with the fluid the integral vanishes, and, 
since \^ does not vary along a vortex line, d^/ds'.ds' is the same for PA and AB. 

Thus for both ends of the vortex-tube we have the same value of wo-, since we have 
to take ds at the two sections to correspond to the same dt. Along a vortex-tube, 
therefore, wo- is constant ; and the tube must either be endless or have its ends on the 
free surface of the fluid, if a free surface exist. In both cases it is really endless. 

21. [February 12, 1909.] The equations of acceleration for any two directions 
drawn from the point P are by § 2 

ou du _ 3V _ 1 dp \ 

dt ds 3./: p 3a; 

\ (35), 

dv dv = __8V_ 1_ dp_ 
dt % ~dj p" dy J 

where dx, dy are infinitesimal steps, and u, v the velocities, in these two directions. 



14 PROFESSOR ANDREW GRAY, 

For simplicity we take the two directions at right angles to one another. If we 
differentiate the first of these with respect to y and the second with respect to x, and 
subtract the first result from the second, we obtain 

i +@ * = % + ^ + % = % + V + %T (36) ' 

where £, >/, £ are the components of spin at P about axes perpendicular to the three 
rectangular planes yz, zx, xy ; u, v, w the component velocities at P ; and the 
divergence. Two similar equations hold for £, >?, and can be written down by 
symmetry. 

These equations are equivalent to those given by v. Helmholtz, as generalised 
for the case of a fluid of non-uniform density. By writing 

©=-±^ 
p dt 

the equations can be put in the more usual form, and in another, from the second 
expression on the right of (36). 

Now let £ be the resultant spin at P, that is, let the axis of z be taken along the 
direction of spin. Then at P, £= n = 0, and we obtain, writing w for £, 

— = - — + - oj . . . . (37). 

dt \dx dyj K ' 

But du/dx + dv/dy may be regarded as implicitly a function of t which changes in 
value as the particle of fluid followed by the total differentiation moves, and in general 
will have a positive or negative yjm'te value. If we suppose the axis of z to turn so as 
to be always in the direction of the axis of spin, then, generally, we shall have at 
each instant dwjdt = — ko, where k is a finite multiplier. Integrating then for a 
moving element, we obtain 

log a) = ct + c 

or .... (38), 

o) = e c 'e CT 

where t is the time interval of integration, and c is a proper mean value of 
— (du/dx + dv/dy) for each successive instant of time, and c' is a constant. If now 
we suppose that initially w is very small, or zero, c' must be a very large, or infinite, 
negative constant. Thus w, if at first imperceptible, would not, in any finite time, 
ac(juire a perceptible value. We obtain again, therefore, the theorem of the per- 
manence of the non-vorticity of a portion of a perfect fluid. 
22. Returning now to (36), write in it 

dt dx dy dz dt 

and multiply both sides of (36) by dxdydz, and integrate for the space contained 



NOTES ON HYDRODYNAMICS, CHIEFLY ON VORTEX-MOTION. 15 

within a fixed closed surface. From the first form of the right-hand side of (36) 
we obtain, since 

dx dy dz 

the result 

I I lj^dxdydz= I l(Zq n -wo> n )dS . . . (39), 

where q n is the normal velocity inwards along the normal at the element rfS of the 
enclosing surface, and w n is the angular velocity about that normal. If the surface is 
taken large enough to include all the vortices, £ and &>„ are zero at every element of the 
surface, and we have 

f f f d ^dxdydz=0 .... (40). 

A similar result is obtained, of course, for each of the other components. The 
volume-integral of the time-rate of change of each component of angular velocity 
of spin has thus the value indicated by the right-hand side of (39), and vanishes for 
any surface enclosing all the vortices. 

If we take the second form of the right-hand side of (36) we get 

fffKdzdydz= (Uq n dS - f f nmq cos ^dS - f j f (v d J + vp- + vB^dxdydz . (41), 

where n is the 2-direction cosine of the inward-drawn normal to ds, and <p is the 
angle between the directions of q and (p at t/S. By (40) this gives the equation 

f ((ww,, - nwg cos </>)r/S = f f l(JL + v ^L+w d ^-\dxdydz . . (42). 



( 17 ) 



II. — The Glacial Deposits of Western Carnarvonshire. By T. J. Jehu, M.D. (Edin.), 
M.A. (Camb.), F.G.S., Lecturer in Geology at the University of St Andrews. 
(With Four Plates and One Map.) 

(MS. received January 12, 1909. Read March 15, 1909. Issued separately May 3, 1909.) 





CONTENTS. 




PAGE 






17 


III. Th 


II. The Coast Sections 


19 


IV. Su 




I. Introduction 



PAGE 

III. The Deipt of the Interior ... 38 

IV. Summary and General Conclusions . 45 



Since the time when Agassiz and Buckland made known the former presence of 
glaciers in the mountain valleys of North Wales, much has been written concerning the 
glaciation of Snowdonia, but comparatively little attention has been given to that of 
Western Carnarvonshire. This part of the country stretches south-westwards as the 
broad promontory of Lleyn between Carnarvon Bay and Cardigan Bay. Western Carnar- 
vonshire for the most part lay outside the paths followed by the native glaciers. None of 
the larger valleys of Snowdonia trend in this direction, and so the marks of recent glacia- 
tion are not so fresh and striking in this region as they are further east. Hitherto no 
one has attempted to give any detailed account of the Drift deposits over the whole of 
Lleyn, but various references to the glaciation of the region are found scattered in the 
literature dealing with the geology of Wales. Among these one of the most interesting 
and most accurate is that in Joshua Trimmer's Practical Geology and Mineralogy 
(1841), where he writes of the " Diluvial Deposits of North Wales " (p. 398). Concerning 
Lleyn he says : " The promontory which divides the bays of Cardigan and Carnarvon is 
covered with diluvial deposits of variable thickness, but frequently exceeding 100 feet. 
They consist of sand, clay, and gravel, changing rapidly from one to the other, through 
which are dispersed broken marine shells, and, amidst much local detritus, fragments 
derived from the north and west. Granite is rare, but they afford fragments of Antrim 
chalk, often very slightly abraded. . . . The south-western termination of this 
promontory affords several instances of the accumulation of the transported matter with 
shells, covering masses of angular detritus of the rock immediately subjacent, which 
may be supposed to have accumulated slowly under the atmosphere, after the manner 
of a talus, or to have been collected from the weathered surface of a bare rock by the 
first rush of the diluvial currents." This extract proves that Trimmer had detected 
the "rock rubble" or "head" which underlies the Drift deposits at the western end 
of the promontory. 

TRANS. ROY. SOC EDIN., VOL. XLVII. PART I. (NO. 2). 3 



18 DR T. J. JEHU ON 

Sir Andrew Ramsay, in his Survey Memoir, "The Geology of North Wales" 
(2nd ed., 1881), makes only occasional references to the glaciation of Lleyn. Describing 
the district, he says : " Like Anglesey it partly consists of metamorphic rocks, which 
are supposed to be of Cambrian age, and of Lower Silurian strata, through which 
numerous bosses of rock protrude of a syenitic and dioritic character. Some of these 
are of great extent, and in Bwlch-mawr and Moel Penllech, Yr Eifl or The Rivals, Cam 
Bodvean, Mynydd Nevin, Cam Madryn, Mynydd Mynytho, Mynydd-y-Rhiw, and other 
minor heights, they generally rise high above the plain of glacial drift that shrouds so 
much of the country, and thus form the loftiest hills of the district. The stratified 
rocks, indeed, generally lie in ground so low, and, except on coast cliffs, they are mostly 
so obscured by sands and marine boulder clays, that it is difficult to make out the 
details of this stratification" (p. 206). Again he states (p. 221) that "from the heights 
of Yr Eifl the promontory of Lleyn, but for the chain of Puys, looks like a mere plain." 
He mentions "an angular breccia of post-Tertiary date resting on the slaty strata" at 
Porth Ceiriad and underlying the Drift deposits (p. 208). Beds of angular breccia are 
also noted as occurring in Aberdaron Bay (p. 211). With regard to Bardsey Island he 
observes that " like the mainland the country has been moulded by ice during the Glacial 
period, but the mammillated rochcs moutonnees have since been roughened by weather" 
(p. 212). He describes the ice-sheet which overspread Anglesey as continuing "its 
onward course to some unknown distance, for on the ice-ground rocks above the sea- 
cliff of Trwyn-y-tal, a mile north of the seaward flank of Yr Eifl, well-marked striations 
are seen pointing towards the Straits 43° north of east " (p. 273). 

In Ramsay's Physical Geology and Geography of Great Britain (6th ed., 1894, 
p. 246), mention is made of the occurrence of shell-bearing deposits in the district of 
Lleyn. It is also stated (p. 248) that " at various levels on the low ground between 
Carnarvon and Criccieth there are extensive deposits of sand and gravel, well stratified, 
and much resembling those of Moel Tryfan, but apparently without shells." These 
are overlain by boulder beds. He held the view that there had been two epochs of 
glaciation in Wales separated by an epoch of submergence during which the land was 
drowned to a depth of 1400 feet or more. 

Mr Carvill Lewis in The Glacial Geology of Great Britain and Ireland, 1894 
(p. 355), states that " deep till, with many large boulders, covers the land, and makes the 
moorland extending from Afonwen to Carnarvon." He also noted the peat at Ynysand 
Pant-glas and the " hills of till " at Brynkir ; the " very large beds of coarsely stratified 
drift, full of large boulders and knob-like in shape," occurring near Penygroes. " All this," 
he concludes, "probably belongs to the great ice-sheet, not to the Welsh glaciers." He 
gives a Glacial Map of England and Wales in which the margin of the northern ice is 
depicted as reaching North-Western Lleyn but leaving the southern part of the pro- 
montory free. In common with many other geologists, Mr Carvill Lewis denied that 
the so-called "Intermediate Sands and Gravels" indicate a submergence of 1400 feet. 
He regarded the high-level shell-bearing sands, such as those of Moel Tryfan, simply as 



THE GLACIAL DEPOSITS OE WESTERN CARNARVONSHIRE. 19 

rearranged material derived from the bottom moraine of an ice-sheet which reached 
this elevation. 

Mr T. Mellard Reade, in a paper on " The Drift Beds of the Moel Try fan Area of 
the North Wales Coast," * describes the coast sections from Dinas Dinlle to Nevin and 
also some sections inland. He favours the view that the sands and gravels are marine, 
and that they indicate a submergence of this area in Glacial times to at least 1400 feet. 

The base of the Lleyn promontory lies just outside the mountains of Snowdonia, and 
has a breadth of about 5 miles. South-westwards the breadth varies from 3 to 5 miles, 
becoming reduced to about a mile at the extreme end. None of the great valleys of 
Snowdonia trend towards Lleyn. The stratified rocks of this district are chiefly of 
Lower Silurian (Ordovician) age, and through these protrude a number of igneous masses 
which form an interrupted chain of hills along the whole extent of the promontory. 
Hard grits and shales of Cambrian age constitute the high ground between Porth 
Nigel and St Tudwal's Eoad, and metamorphosed rocks which are now regarded as pre- 
Cambrian extend from Porth-dinlleyn to the end of the promontory and form the Isle 
of Bardsey. These metamorphic rocks rise to heights of 500 feet at the extreme end of 
Lleyn. The lower parts of the district are covered by a mantle of Drift, and if this were 
entirely removed considerable parts of Lleyn would probably be submerged. 

II. The Coast Sections. 

The following account of the superficial deposits begins with the sections at Dinas 
Dinlle and the coast is followed south-westward to the end of Lleyn and thence 
eastwards on the south side of the promontory to Criccieth. 

From Dinas Dinlle south-westwards as far as Gwydir Bay the coast-line is for the 
greater part bounded by cliffs of varying heights which are described on the Geological 
Survey Map (1 inch to the mile) as consisting of stratified clay, sand, and gravel. The 
bed-rock is not seen anywhere along this stretch of coast. 

To the north of Dinas the coast is flat, consisting of alluvial land which extends 
towards the entrance of the Menai Straits. The character of this flat area is shown by 
a boring recently made for water a few yards behind the hotel at Dinas. Eighteen feet of 
grey sand were passed through, in which the shells of Cardium edule were abundant, 
as well as the shells of Scrobicularia piperata, with the two valves often still in apposi- 
tion. This sand is evidently a recent marine deposit. Below the sand a very tough 
bluish-grey boulder clay was reached, and this extends to an unknown depth. 

Dinas Dinlle. — This is a dome-shaped mound surmounted by an ancient encamp- 
ment. One side of the mound is being cut away by the sea, so that good sections, 
which have been described by Mr T. Mellard Reader are exposed. The cliff varies 
in height up to a maximum of about 100 feet. 

* Proc. Liverpool Geol. Soc, vol. vii. (1893), p. 36. 
t Ibid., p. 42. 



20 DR T. J. JEHU ON 

The following succession can be traced in the deposits : — 

4. Soil and Blown Sand. 

3. Upper Stony Clay. 

2. Sands and Gravels. 

1. Lower greyish-blue or purple Boulder Clay. 

The Upper Stony Clay and the Sands and Gravels form almost the entire cliff". The 
Lower Boulder Clay only appears at the base of the cliff at one or two places. This 
Boulder Clay is very tough and homogeneous, being comparatively stoneless. On the 
surface it has a somewhat reddish or purple hue, but when dug into it has more of a 
dark greyish -blue colour. It can at times be traced underneath the pebbly foreshore. 

The Sands and Gravels in this mound generally show distinct bedding, and the beds 
are arched in a curious way and pass under the Upper Stony Till. The gravels and 
sands contain many flakes and fragments of slate and shale. There are some well- 
marked beds of bigger water-worn boulders. The sands vary from very fine yellow to 
coarse grey. The Upper Stony Clay is a true boulder clay, greyish in colour where 
unweathered, but becoming yellowish and friable on weathering. The included boulders 
are of varying sizes, and are scattered irregularly throughout the mass. Many of them 
show well-striated faces. All are subangular or more or less rounded in form and con- 
sist mostly of Welsh rocks, especially greenstones. The large boulders lying on the 
beach below have evidently been derived to a great extent from this Upper Till. In 
places the till tends to become a mass of boulders or coarse gravel. Shells, in a very 
fragmentary condition, occur rarely in the Upper Clay. The only form recognised was a 
Turritella. Chalk-flints are readily detected. Among other erratics found were the 
Ailsa Craig micro-granite, the Goat-Fell (Arran) granite, Dalbeattie granite, schistose 
and serpentinous rocks, possibly from Anglesey, and a reddish felspar porphyry which 
may have come from no great distance. 

The Dinas Dinlle mound is separated by a low cliff* consisting entirely of the Upper 
Till from another mound which occurs a few hundred yards further south in which the 
cliff sections attain heights of 40 to 50 feet. Here the entire cliff is made up of boulder 
gravel showing but little trace of stratification. Opposite the centre of the mound the 
matrix is sandy or gravelly, but towards the two ends it becomes more clayey in 
nature. 

A diagrammatic section of the mound of Dinas Dinlle is given (fig. 1). 

For about 2\ miles to the south-west of the sections just described the coast 
is fringed by lowlands, and the sea has thrown up a ridge of shingle just above the 
high-water line. There have been considerable changes along this part of the coast within 
recent times, for half a mile out to sea are found submerged the remains of an ancient 
camp known as Caer-Arianrod, which figures in some of the traditional Welsh stories 
— the Mabinogion. Cliffs formed by Drift deposits begin to appear again at Point Maen 
Dulan. At first the cliff is low, but it gradually rises, reaching a height of about 80 feet 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 21 

half a mile to the north of Clynnog. It is only here and there that good clean ex- 
posures are displayed. Often the cliff face is obscured by slipping or by overgrowth. 
Good exposures are seen just to the north of Clynnog, and again opposite the village. 
These show that the entire cliff in this region is made up of a greyish till weathering 
yellow, and full of boulders of all sizes. The till encloses lenticular patches of rather 
fine gravel. The deposit is unstratified, but the boulders often become more numerous 
towards the top. The beach below is strewn with big boulders which have fallen from 
the cliff. Shell-fragments occur rarely, but those obtained were too fragmentary for 
identification. 

This Stony Clay represents the Upper Boulder Clay, and is here of unknown thickness, 
for the bottom is not reached in the sections. Far- travelled boulders are not uncommon, 
and the following were obtained from the cliff near Clynnog : — A foliated granite identical 
with that forming the eastern margin of the Criffel mass in the south-west of Scotland, 
Dalbeattie granite, the Goat-Fell granite of Arran, the Mull of Galloway granite, chalk- 




so// and 

blown sand 



Up ^~ 



7^ clay Upper Boulder Clay »»*»**»* bJSTclJ ™* aZtj^'td Upper B^Mr Clay 

Fig. 1. — Diagram of the Cliff Section at Dinas Dinlle. 

flints, Carboniferous Limestone showing crinoid stems, and various schistose and serpen- 
tinous rocks which are probably derived from Anglesey. 

To the south-west of Clynnog the cliffs are again much grassed over or obscured by 
talus. The heights vary up to 80 or 90 feet. Where best seen the section shows — 

2. Upper Stony Clay — 10 to 12 feet. 

1. Gravels, bedded in places — 60 to 70 feet. 

The bottom of the Gravels is not reached in the section. A little further west fine 
brownish sand is seen mingled with the gravelly material, which is also finer here. 

At Aberafon, on the left side of the stream, near its mouth, the cliff is over 100 feet 
high. It is mostly grassed over, but at one place it can be seen to consist of well- 
stratified sands and fine gravel. There are some gritty beds, but the finer sands are 
stoneless. In the sands occur pieces of a carbonaceous material resembling coal. 
Above the gravels and sands and forming the top of the cliff is a clay full of boulders 
and rounded stones. Mr Mel lard Reade * states that he saw a red clay at the base of 
the cliff underlying the Sands and Gravels. This was not exposed when the writer 
visited the spot, probably owing to the slipping of material from above and to over- 
growth. In the gravels of this cliff a pebble of the Ailsa Craig micro-granite was 
found. 

* Proc. Liverpool Geol. Soc, vol. vii. (1893), p. 43. 



22 DR T. J. JEHU ON 

Just west of Aberafon the cliff face is again somewhat hidden by overgrowth. But 
it can be seen to be formed of laminated stoneless sands and gravel, with some of the 
Upper Stony Clay on top. Immediately north-east of the wooden pier the cliff attains 
a height of about 90 feet. The upper 20 or 30 feet are made up of the Stony yellowish 
Clay, and the rest of the cliff of very fine stoneless sands. As traced laterally the stone- 
less sands in this upper part tend to pass into gravels and boulder gravel — the sands 
still forming the lower part of the cliff down to its base. 

Along the whole stretch of coast from Point Maen Dulan to the wooden pier at the 
north-east end of Gwydir Bay the stiff bluish-grey Lower Boulder Clay was not seen. 
But in Gwydir Bay, just south-west of the pier, the Lower Boulder Clay appears again 
at the base of the cliff. It is very tough and homogeneous, and its dark bluish-grey 
colour is strongly contrasted in certain lights with the light-coloured yellow deposits 
above. It is exposed all along this bay, and, on account of its unctuous nature has 
been the cause of much land-slipping, resulting in the formation of a sort of under-cliff. 
This under-cliff is 10 to 15 feet in height, and above it is a gentle slope much grassed 
over. At the east end of the bay the Lower Boulder Clay is seen to be overlaid by 
gravelly material and this again by the Upper Till. How much of the Lower Boulder 
Clay lies under the grassed slope it is difficult to ascertain. The base of this clay is not 
reached in the cliff sections ; indeed, the foreshore is underlaid by this tough sticky 
clay, from which the remains of sedges or reeds are often seen protruding. Shell-frag- 
ments occur sparingly in the Lower Boulder Clay, but they were too small for identifica- 
tion. Boulders of Carboniferous Limestone, beautifully glaciated, occur plentifully in 
the Lower Boulder Clay, chalk-flints are not uncommon, and porphyrites from the 
Dalbeattie area were obtained. 

To the west of Gwydir Bay the coast is rock-bound. The cliffs of Trwyn-y-tal are 
composed of sedimentary rocks of Lower Silurian (Ordovician) age. Considerable 
erosion has gone on since the Glacial period. There are no traces of a raised 
beach platform, but the almost perpendicular post-glacial cliffs are truncated above 
by a rounded slope very similar to that described by Messrs Wright and Muff 
on the south coast of Ireland.* Along this part of the coast, as on the south of 
Ireland, the slope above can be seen to be continued on stacks now separated from 
the mainland. 

To the south-west of Trwyn-y-tal there is a remarkable small dry valley which 
opens seawards above the cliff. It evidently marks an old overflow channel. Some 
drift is seen here in the cliffs, but the Lower Boulder Clay appears to be absent. 

Passing south, the coast is bound by the igneous mass of Yr Eifl (The Rivals), and 
other igneous rocks of a similar nature form headlands to within a mile of Nevin. But 
at Porth-y-nant small bays have been eroded in the softer sedimentary rocks which reach 
the coast between the harder igneous masses. In these bays Drift deposits occur and 
form the cliffs, but the exposures are not good. At the north-eastern end of Porth-y- 

* Scientific Froc. Royal Dublin Soc, vol. x. (N.S.), part ii., 1904, p. 296. 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 23 

nant the stiff bluish-grey Lower Boulder Clay is exposed, in which shell-fragments can 
be seen. It forms the lower part of the cliff for about 10 feet above the modern 
beach. This Boulder Clay is followed above .by sands and gravels up to a height of 
about 70 feet. 

Again, to the south-west of Careg-y-llam, opposite Pistyll Church, deposits of a 
similar nature occur, but the cliffs are much grassed over. As seen from a distance, 
these Drift deposits form a well-marked terrace with a somewhat flattened top, border- 
ing the higher land. 

The cliffs of Porth Nevin and Porth-dinlleyn are described on the Geological 
Survey Map (1 inch to the mile) as being composed of sand and boulder clay. Both 
bays have a horseshoe or crescentic shape and open out to the north. They are ex- 
cavated for the most part in Drift deposits overlying green quartzose and chloritic schists, 
but the north-east end of Nevin Bay is carved out of the Lower Silurian (Ordovician) 
sedimentary rocks. As seen from a distance, the cliffs seem to be bounded above by a 
terrace which extends inland to the base of the hills with a very even surface. Along 
the greater part of this north-west coast the superficial deposits have this remarkable 
flat-topped aspect. 

The smaller bay at Porth Nevin has an extent of about a mile. The cliff attains 
heights of over 100 feet, but the sections are much obscured by talus formed of debris 
which has slipped from above, and the cliff face is often grassed over. The cliff for 
the greater part of its length is made up entirely of sands, gravels, and laminated 
silty clays. At the two extremities of the bay the bed-rock is seen at the base, 
and here the sands and gravels rest immediately on the solid rock. Where, as opposite 
the middle of the bay, the sections are deeper and the bottom of the Drift deposits not 
seen, the typical stiff dark greyish-blue Lower Boulder Clay appears underneath the sands 
and gravels. This clay contains shell-fragments. Many of the included boulders, 
amongst which the commonest consist of Carboniferous Limestone, are remarkably well 
glaciated. 

Where the bed-rock is visible, the sands and gravels are at some places seen to be 
separated from it by broken-up or shattered rock material, a foot or more deep. 
Fragments of marine shells are plentiful in the sands, but these are generally too 
imperfect for identification. This shelly sand is often stoneless, but at places it 
includes irregular masses, pockets, and seams of coarser gravelly material. Sometimes 
the gravel is so coarse that it would be more accurately described as a boulder gravel, 
and this occasionally swells out to form considerable portions of the cliff. The sands 
are stratified, the gravels sometimes show bedding, but occasionally are tumultuous, 
showing no definite arrangement. Contortions can here and there be observed in the 
sandy beds. 

Above the sands and gravels are buff-coloured laminated silty clays free from stones. 
These at places form the upper part of the cliff, but elsewhere in the same bay the cliff is 
capped by gravel or by a brownish-yellow weathered clayey rubble. 



24 DR T. J. JEHTJ ON 

The general succession of the Drift deposits in this bay is therefore as follows : — 

5. Soil. 

4. Yellowish-brown weathered rubbly clay often becoming pebbly or passing into a 

mass of boulders. 
3. Stoneless laminated silty clay. 
2. Sands and Gravels. 
1. Lower Boulder Clay. 

The silty clay (3) shows fine laminations which are arranged in ribands of lighter 
and darker colours, giving a striped or banded appearance as seen from the shore. The 
bands are sometimes wavy, and examples of faulting occur as described by Mr Mellard 
Reade, # and ascribed by him to contraction and loss of volume, either by actual drying 
or by the draining away of their water. 

At the north-east end of the bay the upper part of the cliff consists of Rubble Drift. 
The lower part is grassed over, but appears to be sandy and gravelly. 

Penrhyn Nevin, the promontory separating this bay from Porth-dinlleyn, has rock at 
the base and Drift deposits above. Ice-scratches running N.N.E. to S.S.W. are seen on 
the rocky platform. This platform runs at a height of 1 to 1 2 feet above the modern 
beach, and is now much cut into, being in process of demolition by the waves. In 
appearance it simulates an old shore platform. It is often clear of drift for 7 to 8 yards 
from its seaward margin, and has a gentle slope outwards. Portions of it can be traced 
on both sides of the promontory as well as round its point. Where the rock-shelf passes 
under the cliff of Drift it has often a hummocky and glaciated aspect, in striking contrast 
to the rough and jagged appearance shown by the rocks on the shore below. On the 
Nevin side of the promontory the Sands and Gravels are, as already mentioned, sometimes 
separated from the platform by a foot or so of rock rubble. On the Porth-dinlleyn side 
clays, somewhat laminated, rest on the rock and are followed above by sand and gravel. 

Porth-dinlleyn lies to the south-west of Porth Nevin. This bay is wider than that 
at Nevin, having an extent of about 2 miles. Its western horn is formed by a mass 
of coarse serpentinous rock. The cliff has a very uniform height of about 100 feet. 
Its face is much grassed over, so that good sections are not often displayed. The deposits 
are very similar to those met with in Porth Nevin. They consist largely of stoneless 
sands and gravels capped by the Upper Stony Clay. At the north-east end of the bay 
stoneless, stiffish laminated clays form the lower part of the cliff. Nearer the middle of 
the bay these are seen to pass downwards into the typical tough Lower Boulder Clay. 
At a point just east of the middle a ledge of rock appears from under the Drift. The 
surface of this has a glaciated appearance. Beyond the ledge the Lower Boulder Clay 
appears at the base of the cliff with big boulders imbedded in its upper part. A little 
further west, beds of gravel come down to the bottom of the cliff face, and are followed 
above by stoneless sands. The upper portion of the cliff is here hidden by overgrowth. 

* Geol. Mag., 1891, p. 487. 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 25 

In the stoneless sands pockets of gravel occur. The sands and the finer gravels show 
beautiful examples of oblique bedding. These false-bedded sands and gravels can be 
traced for some distance in the lower part of the cliff. The sands are very fine and marine- 
like, and apparently form the greater part of the cliff. Just before reaching the road that 
comes down to the beach the Lower Boulder Clay is again seen at the base. West of 
the road the cliff is still obscured by vegetation, but appears to be made up mostly of 
sand. At this part of the bay there has been much sliding over the unctuous Lower 
Boulder Clay, giving rise to an under-cliff. The under-cliff is made up of the lower stiff 
clay, and above it the cliff slopes away and is grassed over. The upper part consists of 
yellowish sandy clay full of small boulders. The Lower Boulder Clay is here very dark 
in colour, and presents a striking contrast to ail the deposits above. 

Promontory of Porth-dinlleyn. — A rocky shelf can be seen on the eastern side 
appearing from under the Drift. It has a glaciated aspect where the deposits have 
been recently cleared away. The shelf or platform is well seen just before reaching 
the lifeboat house. At its inner (landward) end it is 15 feet above the modern beach 
and slopes gently seawards. The sea is gradually eating into and destroying it. Just 
beyond the lifeboat house the serpentinous rock shows distinct marks of glaciation — 
the grooves and scratches running approximately N.E. and S.W. Near the extremity 
of the promontory, north-east of the flag-staff, at a height of 25 feet above Ordnance 
datum, the rock is beautifully glaciated, showing deep and broad furrows and ice- 
scratchings running about N.E. and S.W. These are the finest glacial markings noted 
by the writer in Lleyn. 

The following far-travelled boulders were found in the Gravel and Sand series at 
Porth Nevin and Porth-dinlleyn : — 

Several granites from the Dalbeattie area ; a granite with reddish idiomorphic 
felspars much resembling the Shap granite ; the Ailsa Craig micro-granite. The shell- 
fragments in the sands were, as a rule, too broken and minute to be identified, but the 
following forms were obtained : — Turritella communis (Lam.), Nucula nucleus (L.), 
Pecten (Chlamys) islandicus ? (Mull), Astarte sulcata (Da Costa), Cyprina islandica 
(L.), Cardium edule (L.). 

On the west side of the promontory lies the little bay which bears the name of 
Porth Wen. Here the Drift is again exposed in sections which are 40 to 50 feet high. 
The stiff bluish-grey Lower Boulder Clay forms the base of the cliff, and is succeeded 
above by fine yellow marine-like sands and some gravel. 

From this bay to Porth-ysgadan the coast is rock-bound. Some Drift caps the rocky 
cliffs and is usually overgrown with grass. 

Porth-ysgadan. — Fairly good sections of the Drift deposits are seen at this place. 
Opposite the middle of the bay the typical Lower Boulder Clay with included shell - 
fragments forms the lower part of the cliff, and the bottom of this deposit is not 
reached in the section. Above, the face of the cliff slopes away and is grassed over, 
but there are indications that sand overlies the Boulder Clay. The upper part consists of 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 2). 4 



26 DR T. J. JEHU ON 

yellowish weathered clay becoming sandy in places. At the two ends of the bay the cliffs 
are rocky, though sand is seen resting above on the rock surface at some places, while at 
others the rock is covered by fine rubbly material with yellowish weathered clay above. 

Samples of granites from the Dalbeattie area were picked up on the beach below. 
These had undoubtedly come from the Drift in the cliff above. 

Amongst the shells included in the Lower Boulder Clay the following were recog- 
nised : — Purpura lapillus (L.), Astarte sulcata (Da Costa), Astarte (Nicania) compressa 
(Mort.), Gyprina islandica ? (L.), Corhula gibba (Olivi). 

All along this part of the coast the rock surfaces when visible have a mammillated 
and smoothed appearance where they pass under the Drift deposits. 

Immediately to the south-west of Porth-ysgadan lies a small bay opening westwards 
to the sea, which appears to be nameless. The two arms of the bay are rocky, with 
grassy slopes above. Midway between the arms the Drift comes down to the beach in 
such a way as to suggest that we have here the buried channel of an old stream, or what 
was once an inlet of the sea. The cliff of Drift reaches a height of about 50 feet. It seems 
to consist entirely of the Lower Boulder Clay in which shell-fragments are again common. 
Towards the surface the clay is weathered and yellowish in colour, but there is no 
reason for separating this from the Lower Boulder Clay. Where, as in this bay, the Sands 
and Gravels are wanting, it is sometimes difficult to know whether the yellow clay 
which usually forms the upper part of the cliff is the weathered portion of the Lower 
Clay or the representative of the Upper Boulder Clay. Where typically developed, 
however, the Upper Boulder Clay is less homogeneous, somewhat more sandy, and contains 
more boulders than the Lower Boulder Clay. 

In this small bay, at the two sides, the Lower Boulder Clay appears to be separated 
from the bed-rock by a little rock rubble. 

Porth-y-givylan lies half a mile further to the south-west and opens westwards. 

It is separated on the north side from the open sea by a ridge or projection of bare rock 

running more or less parallel with the coast-line. The other or mainland side of the 

bay is also rocky in character. Between these rocky sides a mass of Drift forms the 

entire cliff and rests immediately on the modern beach, and it is out of this material 

that the bay is still being eroded. The Drift consists of tough bluish-grey Boulder Clay 

at the base, with included shell-fragments, followed above by a boulder gravel with a 

somewhat clayey matrix, and this in turn is succeeded by a yellowish clay. So we have 

here : — 

3. Yellowish weathered Upper Boulder Clay. 

2. Boulder Gravel. 

1. Lower Boulder Clay. 

These deposits appear to have filled up what was either the channel of a stream 
or a narrow inlet of the sea. At the present day a small stream is seen on the 
north-east side of this Drift-filled depression which is cutting out for itself a new channel 
in the solid rock and coming clown to the beach in a waterfall to escape to sea by a 



Trans. Roy. Soc. Edin. 



Vol. XLVII. 



Dr T. J. Jehu on '"The Glacial Deposits of Western Carnarvonshire." 




2 



To face p. 26. 



^^ 



■Wh^" 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 27 

recently eroded channel at the base of the rocky ridge which bounds the bay on its sea- 
ward side. 

Following the coast-line south-westwards we come next to Porth Uchain, a very 
small inlet with rocky cliffs at the sides attaining a height of about 40 feet and capped 
by yellowish weathered clay. Opposite the middle of the bay the Lower Boulder Clay 
appears and forms the lower part of the cliff; above the clay there is a slope which is 
grassed over. 

From this point to the next bay — Porth Caiman — the cliffs are rocky and have a 
steep slope to the sea. The bay which follows is broad, and the northern portion of it 
is sometimes known as Porth Penllech. Here the cliff attains heights of 30 to 40 feet. 
In places the lower part of the cliff is rocky, but at one spot rather stiff dark grey 
boulder clay with many included stones is seen. This becomes yellowish and friable 
where weathered. It probably represents the Upper Boulder Clay, for a little further on 
it is seen to overlie gravel and boulder gravel. 

The remaining portion of the bay is usually called Porth Colman. At its southern 
end the base of the cliff is formed by rock in situ for heights varying up to 18 or 20 
feet. Along a part of this shore a rock-shelf or platform is seen, reminding one again 
of the raised shore platform described by Messrs Wright and Muff as occurring on 
the south coast of Ireland (fig. 2). The platform lies 10 or 12 feet above the modern 
beach and slopes gently upwards to about 15 feet, where it disappears under the Drift 
deposits. It has an exposed breadth of 6 yards or so in places. The surface of the 
platform is covered with lichens or partly grassed over. At its inner end it is covered 
by bluish-grey Boulder Clay containing shell-fragments. This is succeeded above by 
the Upper Yellowish Stony Clay. As we pass towards the middle of the bay the plat- 
form disappears and the cliffs show steep rocky faces with a capping of Drift forming a 
gentle slope landwards. Near the middle of the bay is a remarkable little inlet with 
steep canon-like sides from which a small stream issues. A few yards up this inlet 
the stream is seen to descend by a waterfall over rock to the level of the beach. 

The following erratics were obtained from the Boulder Clay at Porth Colman : — 
Several granite boulders from the Dalbeattie area ; a diorite very similar to that of 
Col vend, Dalbeattie ; boulders of hornblende porphy rites, probably from the Dalbeattie 
area, and chalk-flints. 

The following species were identified from amongst the shell-remains found in the 
Boulder Clay at Porth Colman: — Turritella communis (Lam.), Astarte sulcata (Da 
Costa), Cyprina islandica ( L.), Venus ( Ventricola) Casina (L.), Cardium echinatum (L.). 

Around Penrhyn Colman the coast is rocky, and at Porth-llefesig the rocky cliffs 
reach heights of 50 to 60 feet. Above the rock lies a grassy slope of Drift. At the 
northern end of the bay boulder gravel lies immediately above the rocky cliff, and this 
is capped by stiff clay rather full of boulders and weathering yellow near the surface. 
This clay contains some shell-fragments, but the only one which could be identified was 
Astarte sulcata, 



28 DR T. J. JEHU ON 

Porth Whiting. — For the greater part of its extent the cliffs are rocky, but about 
the middle of the bay rock rubble is seen forming the base of the cliff where rock in 
situ is wanting. Above the rubble is a slope of grass which covers what appears to be 
a clay weathering yellow. At the south-western horn of this bay ice- scratches with a 
N.E. to S.W. direction were noted on the rock surface just where it disappears under 
the Drift. Between Porth Whiting and Porth Ferin some well-glaciated rock surfaces 
are displayed, the vertical face of the rock being sometimes grooved and polished in a 
direction parallel to the shore-line. Above the rock, fine yellow sand is at places 
exposed from under the grassy slope, and this is overlaid by greyish Boulder Clay 
weathering yellow. Further south this stiff Boulder Clay rests immediately on the rock 
surface, and is even sometimes overlaid by yellow marine-like sand. So it may repre- 
sent the Lower Boulder Clay. 

At Porth Ferin an artificial cutting has been made through the rock at the top of the 
cliff. Stiff dark blue boulder clay, which is apparently the Lower Boulder Clay, is seen 
overlying the rock. It is comparatively stoneless, though some well-glaciated boulders 
were found in it. 

No sections of the Drift are again seen until we come to Porth Oer. At the north- 
east end of this bay is solid rock, and on this, at a height of 15 to 18 feet, are beds of 
sand and gravel more or less concreted and much resembling raised-beach material. 
Just south of this a roadway runs down to the beach, and on the southern side of it the 
rocky side of the bay slopes down to the beach and is followed by rock rubble or 
" Head " which in turn is seen to pass under typical dark greyish-blue Boulder Clay. 
This is in all probability the Lower Boulder Clay, and it forms the entire cliff opposite 
the middle of the bay to a height approaching 100 feet. Included in this clay are 
irregular masses of marine -like sand and here and there a little gravel. Shell-fragments 
occur in the Boulder Clay, and the following forms were obtained and identified : — 
Pectunculus glycimeris (L.), Astarte sulcata (Da Costa), Cyprina islandica (L.), Turri- 
tella communis (L.). 

At the south-west horn of the bay a bench of rock is again seen 10 to 15 feet above 
the modern beach. It is much like that seen at Porth Colman, and suggests the possi- 
bility of the presence of an old shore platform here. Stiff* bluish-grey clay (Lower 
Boulder Clay) rests on it, but at places the clay is separated from the rock by coarse 
rubbly material about a foot deep. 

No exposures of Drift deposits are seen from Porth Oer to the end of the promontory 
of Lleyn. The coast is rock-bound with a steep slope, the greater part of which is 
grassed over. 

Lying to the south-west of the mainland and at a distance of 2|- miles from it is the 
Isle of Bardsey. It is composed entirely of metamorphosed rocks, probably of pre- 
Cambrian age. The island has an extreme length of about 1 f miles, and its greatest 
breadth is three-quarters of a mile. The highest portion of the isle lies in the north-east, 
facing the mainland. There, from an elevation of about 500 feet, the land slopes away 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 29 

steeply to Bardsey Sound. The south-west side of the island consists of lowland rising 
gently towards the higher ground. This lowland is covered by a skin of Drift. The 
whole island has the appearance of having been moulded by ice during the Glacial period, 
but, as Ramsay observes,* "the mammillated roches moutonnees have since been 
roughened by weather." The island as a whole may be regarded as an example of the 
phenomenon known as " crag and tail," the crag facing the north-east, from which direc- 
tion the ice-sheet came. 

At Porth Solfach, on the island, the low cliff is composed of yellowish-brown clay 
showing a depth of 6 to 7 feet. Though weathered near the surface, the clay gets 
stiffer below. Bits of shells in the clay were too fragmentary for identification. 
Amongst the included boulders were found chalk-rlints, the Ailsa Craig micro-granite, 
Dalbeattie granite, a picrite, a lava closely resembling that of Borrowdale in Cumberland, 
Millstone Grit, Carboniferous Limestone showing crinoid stems. Some of the boulders, 
especially those of the limestone, were well glaciated. 

Again at Henllyn beach, in Bardsey, pebbles were picked up of Carboniferous Lime- 
stone, chalk-flints, and of a reddish-streaked rhyolite. These were probably derived 
from the Drift of the low cliff adjoining. 

Returning to the mainland, the coast is rocky as far as Aberdaron, but the cliffs are 
interrupted by an inlet at Porth Mendwy, where a stream comes down to the sea. Some 
river-like gravels and sands are seen at the sides of this inlet. 

At Porth Pistyll, 1^ miles to the south-west of Aberdaron, the cliff is steep 
and rocky, but above, at a height of about 100 feet, bluish-grey boulder clay of ex- 
cessive toughness was noted — the deposit attaining a thickness of 4 to 5 feet. 
The exposure is a very small one. This stiff clay is capped by weathered yellowish 
rubbly clay. 

Aberdaron Bay. — Situated on the southern side of the extremity of Lleyn, this bay 
is excavated in beds of Bala age let down by faults into the hard schistose rocks of pre- 
Cambrian age which bound them on either side. The sides of the bay are rock-bound, 
but the cliff between the two arms consists entirely of Drift deposits and is described on 
the Geological Survey Map (1 inch to the mile) as composed of sand, clay, and gravel. 
The bay is over a mile broad, but the cliff is interrupted opposite the village where a 
stream debouches. The height reached by the cliff varies considerably up to elevations 
of over 100 feet. At the south-western end of the bay the rocky cliff descends with a 
steep slope to the beach and is followed by rough rock rubble or " Head." This 
" Head " is overlain by more rubbly material mingled with Boulder Clay, the matrix, 
however, becoming sandy or gritty in places. To the west of the wooden pier the 
" Head" is well defined for 8 or 10 feet from the base of the cliff. It dips eastwards at 
a high angle and disappears under the beach level. The material above this well-defined 
" Head " is still rubbly, with big angular boulders of the neighbouring rock, but it 
becomes more and more clayey as it is traced up the cliff face, and the included boulders 

* Survey Memoir, " The Geology of North Wales," 2nd ed., 1881, p. 212. 



30 DR T. J. JEHU ON 

tend to become smaller and less angular. The material forming the upper part of the 
cliff at this place is very similar to the Upper Boulder Clay which is seen at other parts 
of the coast. 

Between the pier and the village the cliff is over 100 feet in height and shows steep 
faces. It appears to be made up almost entirely of the Lower Boulder Clay, which 
shows the usual characteristics and in which shell-fragments are common. The face of 
the cliff is furrowed by running water, and at the top the weathering gives rise to 
pinnacles. Traces of lamination are seen here and there, and masses of fine yellowish 
marine-like sands occur as inclusions in the otherwise stiff homogeneous clay. 
One of these irregular masses of sand reminds one of the great inclusion of chalk seen 
in the cliff sections of the Boulder Clay in the neighbourhood of Cromer. Gravelly 
patches also occur but are less common. 

To the north-east of the village the cliffs show deposits of a more varied character, 
but a careful examination has led the writer to conclude that the whole belongs to the 
horizon of the Lower Boulder Clay. The greater part of the cliff, which attains heights 
of about 100 feet, consists of the typical Lower Boulder Clay. This invariably forms 
the higher portion of the cliff, and weathers into pinnacles with steep sides. Lower 
down and towards the bottom of the cliff the deposit becomes more of a boulder gravel 
or gravel and sand. Midway between the village and the eastern end of the bay the 
section is as follows : — 

5. Stiff greyish-blue Boulder Clay, comparatively stoneless, and containing shell- 
fragments — 40 to 50 feet. 

4. Boulder Gravel — 10 feet. 

3. Dark band of stiff clay — 2 feet. 

2. A talus slope probably covering sand and fine gravel — 10 feet. 

1. Stiff grey Boulder Clay — 2 feet. The bottom of this is not reached in the 
section. 

From a little distance the boulder gravels and sands can be traced as beds running 
across the face of the cliff. The gravel beds appear to form gentle undulations as 
shown in the diagram (fig. 3). Further east slipping has taken place, due to the 
presence of sand in the clay, and there is a tendency to form an under-cliff. At the 
extreme east end there is a fine example of " Head," lying between the Boulder Clay 
and the solid rock. The " Head " consists of angular and flaky fragments showing rude 
bedding, and these fragments are derived from the neighbouring rocks. The section at 
this place is shown diagrammatically in the following figure (fig. 4) : — 

A cave has been excavated in the rubbly material by the waves. The Boulder Clay 
overlying the " Head " contains pockets of boulder gravel. 

Imbedded in the beach at this end of the bay just under the rock cliff is a big 
boulder measuring 66 inches by 24 inches. It is a granite with large idiomorphic 
reddish felspars and closely resembles Shap granite. The felspar is orthoclase ; one 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 



31 



of the crystals measured 2^ inches in length. Several pebbles of the Ailsa Craig rock 
were also picked up on the beach. 




Length of section about 450 feet. 



a — a. Boulder gravel bed. 

b. Laminated clay. 

c. Tough boulder clay weathering into pinnacles. 

Fig. 3. — Diagrammatic Section of the Cliff East of Aberdaron. 



Among the included stones picked out of the Boulder Clay were chalk-flints, the 
Ailsa Craig micro-granite, granite from the Dalbeattie area, the Goat-Fell granite of 
Arran, and the Loch Dee granite of the south-west of Scotland. In addition to these 



Boulder clay 
pebbly in places. 



Boulder clay. 




60 feet. 



' 40 feet. 



Fig. 4. — Section (diagrammatic) East End of Aberdaron Bay. 

there were boulders of Carboniferous Limestone, probably from Anglesey, of a reddish- 
streaked rhyolite, of jasper, and serpentine. 

Amongst the shell-fragments obtained the following have been recognised : — 



Gasteropoda. 

Turritella communis, Lam. 
Buccinum undatum, L. 
Trophon (Boreotrophon) clathratus, L. 
Bela turricula. 



Lambllibbanchia. 

Pectunculus glycimeris, L. 
Mytilus edulis, L. 
Astarte sulcata, Da Costa. 
Astarte (Nicania) compressa, Mont. 
Cyprina islandica, L. 
Tellina (Macoma) balthica, L. 
Cardium echinatum, L. 
Cardium edule, L. 
My a truncata, L. 

A portion of a Belemnite was also found in the Boulder Clay. 

Forth Cadlan ( Ysgo). — The Lower Boulder Clay is not exposed in this bay. About 



32 DR T. J. JEHU ON 

midway between its two sides <i slope of Drift is seen reaching heights approaching 100 
feet. It consists of yellowish weathered clay with large boulders, and contains some 
shell-fragments. Pebbly and sandy layers are seen here and there. At the sides of 
the bay this Drift is separated from the solid rock by angular rubble or " Head." Some 
of the big boulders in the cliff are well glaciated, and the beach below is strewn with 
blocks derived from above. 

Porth Nigel or Hell's Mouth. — Situated on the southern side of the Lleyn promon- 
tory, this bay opens out to the south-west. Its south-eastern side is bounded by the 
high ground of Mynydd Cilan, which is made up of highly resisting hard gritty sand- 
stones and shales of Cambrian age ; the north-western side partly by indurated shales 
of Ordovician age, but chiefly by the greenstones which extend from Mynydd-y-Rhiw to 
Penarfynydd. Between these sides the head of the bay is formed entirely of glacial 
deposits and has an extent of about 4 miles, giving one of the finest sections of the Drift 
to be seen in North Wales. These deposits stretch inland, covering the whole country 
from Llanengan to Llangian and on north-westwards by Llandegwning to Bottwnog, 
thence south-westwards as far as Mynydd-y-Ehiw. Erosion is still going on to a marked 
extent at the head of the h&y. On the Geological Survey Map (l inch to the mile) it 
is stated that the cliff consists of " stratified clays, sands, and gravel from 10 to 40 feet 
high." This description requires considerable modification. At places the cliff attains 
much greater elevations. The general succession is as follows : — 

4. Soil and Blown Sand. 

3. Upper yellowish weathered Boulder Clay, becoming sandy and pebbly in places. 
2. Intermediate Sands and Gravels. 

1. Lower Boulder Clay — massive, tough, and homogeneous. (This Lower Boulder 
Clay descends below the level of the modern beach to an unknown depth.) 

The cliff is highest towards the north-west end of the bay, where it attains elevations 
approaching 100 feet, and shows almost vertical faces which present a striking appearance. 
Here it consists almost entirely of the Lower Boulder Clay, which, indeed, is by 
far the most prominent member of the various deposits exposed in this bay. The 
Intermediate Sands and Gravels are not seen at the western end, but are well displayed 
when we pass towards the eastern end. Taking the deposits from below upwards, we 
have at the base — 

The Lower Boulder Clay. — This is a stiff, homogeneous clay, grey in colour on 
the face of the cliff but becoming darker below the surface. It presents a massive 
appearance, giving almost vertical faces which are traversed by fissures resembling 
joints running from above downwards. These cracks are widened by the action of 
rain and frost, and large blocks are thus wedged off and fall in masses on to the beach 
below. The base of the cliff is reached by the waves at high spring tides, and the 
clay is scooped out, leaving a smooth hollow over which the cliff above hangs. This 
overhanging of the cliff shows the tough nature of the clay. Here and there traces of 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 33 

a kind of bedding or lamination can be seen which may possibly be due to a shearing 
movement, for the clay has all the appearance of having been subjected to great 
pressure. Boulders are fairly common but not abundant. They are distributed pell- 
mell throughout the clay, not in clusters or patches but singly. The majority are about 
the size of one's fist, and are subangular and beautifully glaciated. Far-travelled rocks 
are common, and include chalk-flints, various south of Scotland granites, especially from 
the Dalbeattie area, the Ailsa Craig micro-granite, a granite with reddish felspars, 
Carboniferous Limestone, Millstone Grit, etc. Pebbles of the Ailsa Craig rock and of 
the other Scottish granites were also picked up on the beach and had evidently been 
derived from the cliff. Bits of wood were also found included in the clay. Marine 
shells occur scattered irregularly in the clay, generally in a very broken condition, but 
some complete shells were found. The forms identified are given below. The fore- 
shore is for the most part covered by pebbles, but at some places it consists of buttery 
dark blue clay evidently due to the running of the Lower Boulder Clay. 

The Intermediate Sands and Gravels. — These are somewhat inconstant in their 
occurrence, but towards the eastern end of the bay they form the greater part of the cliff. 
Usually they are seen to overlie the Lower Boulder Clay, but at places they rest 
immediately on the modern beach, where the Lower Boulder Clay is not exposed. 
Where the sands follow the clay in the cliff there is often much slipping. Sometimes 
this series consists entirely of fine siliceous sand, sometimes of fine pebbly material, and 
sometimes of coarse gravels. The sands and gravels often, but not at all places, show 
stratification. These deposits vary much in thickness and are apt to die out when 
traced laterally. At Porth Nigel no contortions were observed in the sandy layers. 
Here and there the sands become largely ferruginous. No shell-fragments were noted. 
The pebbles in the gravels are of rocks very similar in their origin to those found in the 
Lower Boulder Clay. 

The Upper Boulder Clay is not very conspicuous in these sections. Where it occurs it 
sometimes overlies the Intermediate Sands and Gravels, and sometimes rests immediately 
on the Lower Boulder Clay. It is yellowish in colour, and much weathered, and tends 
in places to become sandy and pebbly, or to pass into a rubble Drift. Ice-scratched 
stones can be picked out, but are not so frequently met with as in the Lower Boulder 
Clay. The included boulders contain erratics, but those from a great distance are not so 
common as in the Lower Boulder Clay. Fragments of marine shells are rare, and those 
obtained were very fragmentary. 

At the west end of the bay a little angular local detritus or " Plead " lies on the solid 
rock where it slopes down to the beach. There is much slipping and overgrowth here 
but the Lower Boulder Clay, which forms the base of the cliff of Drift, appears to overlie 
this "Head." Towards this end of the bay the cliff attains heights approaching 100 
feet, and consists from top to bottom of the massive Lower Boulder Clay. Further east 
a somewhat sandy band runs along the face of the cliff 15 feet from the bottom. This 
band has a thickness of 2 to 3 feet, and when traced laterally it has more of the nature 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I, (NO. 2). 5 



34 DR T. J. JEHU ON 

of a stiff bluish clay which is laminated. This can be traced continuously for a 
considerable distance, but eventually it comes down to the level of the beach and dis- 
appears. As we pass nearer the middle of the bay laminations appear near the top of 
the cliff. The cliff face in this region is not so steep, and considerable slipping has 
taken place so as to form a sort of under-cliff with a rough slope above. The top of the 
cliff is here formed of the Upper Boulder Clay, which is yellowish in colour and some- 
what sandy, and has a thickness of 12 to 20 feet. Where landslips have occurred they 
have given rise to a lamination or stratification in all the deposits, the beds dipping 
inland at a high angle. At one place about 1|- miles from the west end of the bay, a 
bed of peat 1 8 inches thick was observed near the base dipping inland at an angle of 
about 35°. It could be traced for a horizontal distance of 20 yards, and was resting on 
boulder clay and overlain by sandy material and this again by boulder clay. The peat 
contained reeds or sedge and was full of land-shells, such as Helix. It had evidently 
slipped from the surface and been in turn covered by further landslips. 

A little to the east of the middle of the bay there is a good exposure of the Upper 
Boulder Clay. The cliff is here 80 to 90 feet high, and the upper 25 feet consist of 
rubbly clay immediately overlying the Lower Boulder Clay. This dies out when traced 
laterally. Further east fine yellow sand is seen near the top of the cliff. The cliff 
becomes lower as we travel towards Mynydd Cilan, and the Intermediate Sands and 
Gravels become more prominent. Where the succession is most complete we have a 
section as follows : — 

5. Soil and Blown Sand. 

4. Upper Boulder Clay, weathered and yellowish in colour, stiff in places, but often 

becoming sandy and pebbly or even rubbly. 

3. Fine yellow siliceous sand becoming somewhat clayey in its lower part. 

2. Laminated greyish-blue stoneless clay becoming somewhat sandy in places, and 

distinctly marked off from 1 . 

1. The massive Lower Boulder Clay which forms the base of the cliff and the bottom 

of which is not reached. 

Near the extreme east end of the bay the gravels of the Intermediate series come down 
to the base of the cliff, but the Lower Boulder Clay is exposed on the foreshore below. 

At a short distance from this end peat appears in the upper part of the low cliff. 
This peat can only be traced for a horizontal distance of about 1 00 feet. The section 
here is as follows : — 

5. Blown sand — 5 feet. 

4. Peat— 3 1 feet. 

3. Stiff blue peaty clay — 2 feet. 

2. Greyish-blue somewhat sandy clay, pebbly in places — I foot. 

1.. Stiff Boulder Clay with a tendency to lamination and comparatively stoneless. 
(This rests on the beach and its bottom is not reached.) 



i 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 35 

Hazel-nuts were found in the peat, and the bark of birch trees. A little rubble rests 
on the rocky slope at the end of the bay, but its relation to the Lower Boulder Clay is 
obscured by slipping. 

Amongst the shells and fragments of shells gathered from the Lower Boulder Clay 
at Porth Nigel the following have been identified : — 



Gasteropoda. 

Ttcrritella communis, Lam. 

Buccinum undatum, L. 

Tritonofusus gracilis, Da Costa. 

Murex (Ocinebra) erinaceus, L. 

Purpura lapillus, L. 

Trophon {Boreotrophon) dathratus, L. 

Trophon [Boreotrophon) scalar iformis, Gould. 



Lamellibranchia. 

Pectunculus glycimeris, L. 
Mytilus edulis, L. 
Astarte sidcata, Da Costa. 
Astarte (Nicania) compressa, Mont. 
Astarte (I'ridonta) arctica, Gray. 
Cyprina islandica, L. 
Tellina (Macoma) balthica, L. 
Mactra (Spisula) solida, L. 
Venus (Venfricola) casina, L. 
Gard.ium edule, L. 
Mya truncata, L. 
Corbula gibba 1 Olivi. 
Nucidana, sp. 

A cirriped was also found in the ciay. # Complete specimens of Murex erinaceus and 
Tumtella communis were obtained, and complete single values of Astarte sulcata and 
A. compressa. 

Porth Ceiriad is situated at the end of the promontory which separates Porth Nigel 
from St Tudwal's Road. This bay opens to the south, and is carved out of the hard 
gritty sandstones and shales of Cambrian age between Penrhyn-du and Mynydd Cilan. 
It is nearly a mile broad. The Lower Boulder Clay is not exposed in this bay. The 
cliff attains heights of 50 to 60 feet, and slopes away inland at the top. It consists of 
yellowish weathered Drift with boulders which are subangular in shape, and many of these 
show ice-scratches. Beds of pebbly material and sand are seen lying in or below the above, 
and even in places extending to near the top of the cliff. Toward the top, however, 
the Drift is usually rubbly in character. At the two sides where the rocky cliffs come 
in ; the Drift is separated from the bed-rock by a mass of rubble or " Head." Large 
boulders derived from the cliff lie on the beach, and these are chiefly of local Welsh 
rocks, but specimens of the reddish-streaked rhyolite are fairly common. 

St Tudwal's Road. — The only place where the Drift is exposed is at the south- 
western end, where it consists of a rubbly clay, yellowish and weathered. This probably 
belongs to the Upper Boulder Clay horizon. The whole of this bay, with the exception 
of the rocky promontory and the estuary at Abersoch, is bordered by sand dunes. At 
the back of the sand dunes marshy land occurs at some places, as at Cors Llyferin, and 
this is probably underlain by Boulder Clay. The promontory on which a part of 

* For the determination of these and other shells mentioned in this paper the writer is indebted to Mr Henry 
Woods, M.A., Lecturer in Palaeontology at the University of Cambridge. 



36 DR T. J. JEHU ON 

Abersoch village stands is grassed over, but marine-like sands appear wherever the 
ground is dug into, and apparently these sands mantle the land around the village. 
Drift is not again exposed along the coast until we pass east of Mynydd Tir-y-cwmmwd. 
Llanbedrog. — Between Mynydd Tir-y-cwmmwd and Careg-y-defaid, both of which 
are rocky and formed of felspar porphyry, the cliff consists of Drift deposits, and good 
sections are exposed for a distance of over a mile. Stiff dark brownish Boulder Clay is 
seen on the foreshore at the west end of the bay, and at one place this clay forms the 
lower part of the cliff. This is the Lower Boulder Clay, and it underlies the Sand and 
Gravel series which forms the cliff all along the bay. The cliff has a very uniform 
height of a little over 30 feet. The sands and gravels are stratified, and occasionally 
show oblique bedding. The sands are orange-coloured and somewhat ferruginous ; they 
yielded no shell-fragments. So in this bay we have represented 

2. The Intermediate Sands and Gravels. 

1. The Lower Boulder Clay. 

On the beach, boulders of Carboniferous Limestone were seen and of granites from 
the Dalbeattie area ; very common also were boulders and pebbles of the reddish- 
streaked rhyolite which is so often found scattered about Lleyn. 

From Careg-y-defaid to Pwllheli the coast is bounded by sand dunes. An old 
coast-line can be traced some distance inland in the shape of a cliff which is now grassed 
over. This old cliff is separated from the dunes by flat land some of which is marshy. 
Sand dunes again border the coast-line east of Pwllheli, as far as the felspar porphyry 
promontory of Pen-y-chain. 

The greenstone mass of Careg-y-rhimbill opposite Pwllheli is being rapidly de- 
molished by quarrying operations, but its summit presents a moutonneed aspect. 

At Pwllheli, in connection with the Harbour Works, a boring was made in the estuary, 
and the following deposits were traversed : — 

3. Mud — 6 inches. 

2. Ballast and sand — 4 feet. 

1. Stiff blue clay — 8 feet. (The bottom of this clay was not reached.) 

East of Pen-y-chain, cliffs composed of Drift deposits are again met with and are 
being eroded by the sea. The best sections are displayed between Afonwen and Afon 
Dwyfor. 

Afonwen to Afon Dwyfor. — On the Geological Survey Map (1 inch to the mile) 
these cliffs are marked as consisting of " stratified sands, clay, and gravel." The main 
deposits exposed along this stretch of coast are : — 

2. Upper Boulder Clay and Rubble Drift. 
1. Sands and Gravels. 

The Lower Boulder Clay of the Lleyn promontory is not seen here. The sands are 
often well bedded, the gravels generally coarsely bedded. At places the sands show 






Trans. Roy. Soc. Edin. 



Vol. XLVII. 



Dr T. J. Jehu on " The Glacial Deposits of Western Carnarvonshire." 




Fig. 5. — Afonwen Section, near west end. Coarse gravels capped by the 
Upper Boulder Clay. 




above by Boulder Gravel and the Upper Boulder Clay. 



To face p. 36. 



.♦la ^ 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 



37 



beautiful examples of oblique bedding. The deposits in the western half of the section 
are very varied, and change frequently in nature when traced laterally. Near Afon- 
wen a bed of peat is seen under the soil and blown sand, and where the cliff descends 
the peat comes down on to the beach. The well-bedded sands generally underlie the 
coarse gravels. Inclusions or pockets of stiff yellow clay occur in the gravels. At some 
places the pebbles or rounded boulders are heaped in a tumultuous fashion and show no 
traces of bedding (fig. 5). A curious accumulation of well-bedded ferruginous sand 
occurs at one place in the cliff, the beds being arched up (fig. 6). This may repre- 
sent an intra-glacial channel which became filled up with sand arranged in beds. When 



Soil. 



Upper Boulder Clay 
—10 feet. 



Coarse gravels with -ffi 

some sand — 18 feet. —' 



Fine-bedded sand. 



Sand, more clayey 
at the bottom — 15 
feet. 




Bedded sand. 
Beach. 



10 feet = l inch. 
Fig. 8. — Diagram of the Afonwen Section near the East End. 



the walls of ice disappeared the beds would tend to bend over at the sides so as to give 
an archlike arrangement. These ferruginous beds are followed by boulder gravel and 
over this lies stony till full of boulders which are glaciated. 

About midway along the section peat again appears. Two beds can be seen 
separated by a thin band of very stiff bluish-grey clay with rootlets. The lower bed 
of peat is underlain by tough bluish clay of a similar nature. This peat was examined 
by Mr F. L. Lewis of Liverpool University. He states that it yielded : — Viola 
palustris, Carex sp. (fragments), Menyanthes trifoliata seeds (fairly numerous), frag- 
ments of Sphagnum leaves, Potamogeton fruits in fragments (most probably those 
of P. natans), also scraps of birch bark and birch wood. Further east these beds 
of peat seem to unite owing to the dying out of the intermediate band of stiff 



38 DR T. J. JEHU ON 

clay (fig. 7). Where the peat appears the cliff is low and the sands and 
gravels disappear. Further east the cliff rises again to heights of about 50 feet, and 
consists from top to bottom of gravels and bedded sands, the sands forming the lower 
part. As we approach the Afon Dwyfor end the Upper Boulder Clay, which is here 
very stony in character, is seen forming the upper part of the cliff and overlying the 
coarse gravels, which in turn rest on the sands (fig. 8). 

No shell-fragments were seen in any of these deposits, and the pebbles and boulders 
are all apparently of Welsh origin, no foreign erratic having been noted. 

Sections are again exposed on both sides of Criccieth, where the cliffs consist of 
the Upper Boulder Clay. A fine section is shown east of the promenade at Criccieth 
(fig. 9). It is a typical boulder clay containing big boulders which are often beautifully 
striated and polished. The shore is strewn with blocks and boulders derived from the 
cliff. No shell -fragments were noted and no foreign erratics. The boulders consist 
chiefly of greenstones derived from the Snowdonian area. 

III. The Drift of the Interior. 

Away from the coast good sections are not often displayed. The greater part of the 
interior is covered by a mantle of Drift which often rounds off the lower hills, but the 
higher mountains rise above all the glacial deposits. Here and there, even on the 
lower ground, bosses of rock protrude through the Drift, and these often show rounded 
outlines and a more or less glaciated aspect, but the smoothness has generally been 
much destroyed by long exposure to the weathering agents. 

The region between Carnarvon and Afon wen is for most part bleak and dreary. Not 
only is there a deep covering of the superficial deposits over this area, but the Drift 
often forms mounds, ridges, and elongated hills. Many large boulders are still seen 
strewn over the surface, although the land has been to a great extent cleared of the 
loose stones which have been used in building cottages, farmhouses, and the walls and 
dykes which now form a network over the whole country-side. In the hollows and 
lower parts peat occurs, as, for instance, at Pant-glas, where it is being cut to a depth of 
4 feet, and again north of Ynys. 

The undulating plain of Drift rises gradually from the coast-line at Dinas and 
reaches a level of about 400 feet at Penygroes. Between Penygroes and Clynnog Fawr 
lies Y Foel — a hill rising G50 feet above sea-level. At the summit of this hill there is 
an outcrop of slaty rock, but with this exception the entire hill is rounded off with 
Drift. Sections of the superficial deposits can be seen on the banks of streams, in 
railway cuttings, and in gravel-pits on the surface or sides of the moraine-like hills. 
Borings made for water often throw light on the deeper deposits. The mounds and 
hills of Drift which rise above the general surface of the plain are made up of roughly 
bedded coarse grey sand, gravels, and boulder gravel in which some large angular and 
subangular boulders are frequently seen. The materials are for most part of local 



Trans. Hoy. Soc. Ediu 



Vol. XLVII 



Dr T. J. Jehu on " The Glacial Deposits of Western Carnarvonshire." 




Fig. 7. — Section east of Afonwen. Two beds of peat are seen separated by tough 
bluish-grey clay. As traced laterally these two beds of peat unite. 




Fig. 9. — Section of the Upper Boulder Clay In the cliff east of the Marine 
Terrace at Criccietli. 



To face p. 38. 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 39 

origin, but erratics such as far-travelled granites and chalk-flints are occasionally found. 
In these hillocks and ridges the material is generally more or less rounded. 

The Drift of the plain rises on the flanks of the higher hills to heights of about 600 
feet. But patches of glacial accumulations are met with at much greater heights. On 
Y Gyrn Goch, south-west of Clynnog Fawr, Drift occurs up to a height of 1200 feet 
above sea-level. It consists of yellowish clayey sand in which, however, no traces of 
marine shells were found. Mr Mellard Re ade notes the occurrence of "splendidly 
polished quartz-grains in the angular semi-decomposed gravel from near the summit of 
the Pass between The Rivals and Mynydd Carnguwch, about 800 feet above sea-level." # 

As we pass west of a line taken as running from Clynnog Fawr to Pwllheli, the 
moraine-like mounds and esker-like hills and ridges tend to disappear. The gravels and 
sands continue, however, to cover much of the land and to form a mantle over the lower 
hills. The sands tend to become fine, and are more marine-like in aspect. The fiats 
and hollows are underlain by Boulder Clay. The hills in the neighbourhood of Nevin, 
as, for instance, Garn Bodvean, which attains an altitude of over 900 feet above sea-level, 
have a distinctly moutonneed aspect, and were during the Glacial period buried under 
the ice-sheet which moved from north-east to south-west. Exposures of the Drift are 
few in number and very poor in the western part of Lleyn, and boulders which at one 
time dotted the surface have for most part been cleared away. Beginning at the 
eastern end of the area investigated, the following notes were made of such exposures as 
could be found inland. 

Poat-Cryehddivr, nearly a mile south of Llanllyfni. Mr Mellard Reade found a 
fragment of Eskdale granite, a Brachiopod from a Carboniferous shale, and fragments of 
schists probably from Anglesey in typical till exposed at a level of 400 feet above the 
sea on the right-hand side of the stream Afon-ddu, just above the bridge. Sections in 
the neighbourhood of Llanllyfni, made in driving headings in slate-quarrying operations, 
show boulder gravel lying upon hard buff-coloured clayey till. 

Again, Mr Reade found far-travelled granites, one of which was located as from 
Eskdale, in loamy and gravelly material seen in a section 30 to 40 feet high on the 
right bank of the Afonddu, below Pont-Crychddwr, 350 feet above the sea. 

On the right bank of the Avon Llyfni there is a good section 35 feet in height 
showing sandy clay full of boulders and gravel. 

Near the railway station at Penygroes Mr Reade describes t "a large mound of 
Drift, the major axis of which is one-third of a mile long, the direction being approxi- 
mately north and south." The width of this mound is about 200 yards. The mound 
is composed largely of rounded stones, with sand and fine gravel, and stiff clay occurring 
in it irregularly. This can be seen in a gravel-pit by the roadside on the way to Pont- 
y-Cim. Between Pont-y-Cim and Y Foel there lies a circular mound of Drift known as 
Craig -y-ddinas which has a height of about 40 feet above the road. It consists largely 
of boulder gravel and fine gravel often with a clayey matrix. The mound is surmounted 

* Proc. Liverpool Geol. Soc, vol. vii. (1893), p. 45. t Ibid., p. 48. 



40 DR T. J. JEHU ON 

by an ancient encampment. The Foel itself is for the most part covered by a mantle of 
Drift and is grassed over. Several sand and gravel pits can be seen between Clynnog 
Fawr and Llanllyfni ; as, for instance, just off the roadside above Tai'n-lon, and again on 
the side of Y Foel, where the material is gravel and in which a south of Scotland granite 
boulder was picked up. 

At Talarn a new cut for the Afon Llyfni disclosed a section of gravelly drift 440 
feet above the sea. Sections of similar drift can be seen in most of the slate quarries, 
but according to Mr Reade its maximum thickness does not appear to exceed 20 feet in 
the bottom of the valley. This gravelly material has a clayey matrix. 

A gravel-pit of the usual character occurs at Ty-glas, about a mile to the south of 
Tai'n-lon. 

At Ynys-yr-Arch, when visited by the writer, a trench was being dug out to accom- 
modate a churning wheel. The section exposed showed : — 

3. Soil— 1 foot, 
2. Gravel — 4 feet. 

1. Stiff bluish-grey Boulder Clay — 6 feet. (The bottom of this Boulder 
Clay was not reached in the section.) 

Baron Hill, near Bivlcli-y-denvydd, Pant-glas. — Just below the farmhouse on the 
north side boring operations in search of water were being carried out. In the shaft 18 
feet were passed through when the locality was visited by the writer, and the material 
traversed consisted entirely of typical dark greyish Boulder Clay which was weathered 
yellow near the surface. Boulders were common in the deposit and were well glaciated. 
The bottom of this clay was not reached. 

At Llecheiddior Ganol, lying between Pant-glas and Ynys railway station, there 
occurs a gravel and sand pit. The deposits are bedded, and are coarse near the surface 
and fine below. The bottom of the sands is not exposed. No foreign erratics were noted. 

At Plus Llecheiddior, just behind the house, a section of the hillock is disclosed. 
Boulder rubble is seen overlying fine sand which has a depth of 4 feet in the section 
without reaching the bottom. Gravelly and sandy beds of the usual character were 
exposed in a pit at Llecheiddior Mill, but the pit is now filled up. 

In the neighbourhood of Brynhir are many mounds and elongated hills of gravel 
and coarse sand. The included stones are all more or less rounded by water action and 
apparently consist of locally derived or Welsh material. A rough bedding is usually 
shown. The ridges are seen to run for most part from north to south. East of Brynkir, 
at Blain-y-cae, gravel is again seen, but the deposits have become finer, passing into 
sand, as traced downwards from the surface. 

At Grraianog Farm, near Ynys station, there is a small gravel and sand pit just 
;il>ove the house. In this pit a chalk-flint was picked up, and also a granite with reddish 
felspars and black mica, which is undoubtedly a fair-travelled erratic though its parent 
rock- li;is not been located. 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSELRE. 41 

The place-names in this district are interesting and suggestive. "Grainog" is the 
Welsh term for "gravel," and it occurs again and again in this neighbourhood, e.g. 
Bryn-grainog — the gravelly hillock ; Cors Grainog — the gravelly moor ; Ynys Grainog 
— the gravelly isle, etc. Several spots bear the name of Ynys (island), and these 
generally mark hillocks or mounds of gravel and sand standing above low marshy or 
boggy ground which is no doubt underlain by Boulder Clay. Much of this low 
ground is peaty, and in early historic or prehistoric times was probably under water, 
leaving the hillocks standing out as islands. Some of these islands may be mentioned as 
examples : — Ynys-wen, Ynys-y-creian, Ynys-grainog, Ynys-yr-Arcli, Ynys-y-buntan, 
Ynys-hower ; and there are several others. Most of the spots bearing such names mark 
hillocks of sand and gravel, but some denote rocky knolls projecting above the low 
ground, e.g. Ynys-galed — the hard or rocky island. 

Llanaelhiaiarn Gravel Pit. — This lies a quarter of a mile to the north of the 
village of Llanaelhiaiarn, near the mill situated just off the right-hand side of the main 
road from Pwllheli to Carnarvon. It is at a distance of 1^ miles from the sea (Carnarvon 
Bay), and at a height of about 450 feet above sea-level. The section exposed is as 
follows : — 

2. Brownish-yellow weathered clay, somewhat sandy in places, and with subangular 

boulders especially numerous towards the surface — 6 feet. 
1. Stratified sands and gravels — 15 feet. 

The beds of sand and gravel alternate, but towards the bottom of the exposure only 
fine gravelly material and sand are seen. The thickness attained by these gravels and 
sands is unknown, for their base is not reached in the section. Most of the included 
stones and boulders are of local origin, but specimens were also obtained of chalk-flints, 
and of a reddish-streaked rhyolite. 

This pit occurs on the side of a slope which is mostly grassed over. It is 
probable that the gravels and sands cover considerable areas to the north-east of this 
exposure. 

Bryn-givydin. — This place lies on the right-hand side of the road between Chwilog 
and Four Crosses, and near to the latter village. Here the hillock has been dug into 
for sand and gravel. The section is about 12 feet deep, but is much obscured by talus. 
Near the surface the material exposed is earthy, but below are beds of coarse gravel 
alternating with beds of fine gravel. The beds get finer and sandier as traced down- 
wards. Amongst the boulders are several examples of the reddish-streaked rhyolite 
which is so common in Lleyn. Just north of Four Crosses there occur pits showing 
deposits of a similar nature. The sand has usually a greyish colour. No marine shells 
are detected. West of Four Crosses, on the way to Nevin, some pits are again seen 
in fields just off the roadside. In these sands and gravels are again exposed, and these 
are in some of the pits overlain by 4 to 5 feet of bouldery rubble. 

Just east of Penprys, off the same road and at a height of 300 feet above sea-level, 
TRANS. ROY. SOC. EDIN, VOL. XLVII. PART I. (NO. 2). 6 



42 DR T. J. JEHU ON 

there is an exposure showing 8 to 9 feet of coarsely bedded gravelly material which 
becomes finer when traced downwards. Amongst the pebbles were found specimens of the 
reddish-streaked rhyolite and one of a reddish granite which is foreign to North Wales. 
At Tan-y-llyn, near Glasfryn, there is a shaft for water, 30 feet deep, giving a 
good section. The materia] exposed consists entirely of tough, dark greyish-blue 
Boulder Clay full of boulders of all sizes, most of which are beautifully glaciated. The 
clay weathers to a brownish-yellow colour just at the surface. This was the deepest 
section met with in this region. Amongst the boulders the reddish-streaked rhyolite 
was common. 

Another shaft for water was dug at Tyddyn- Ucha about a quarter of a mile east of 
Tan-y-llyn. The depth reached here was 9 feet, and the material traversed consisted 
again entirely of stiff bluish-grey Boulder Clay full of well-glaciated stones. 

At Ty-Cam, a quarter of a mile distant from Four Crosses, in the direction of 
Pwllheli, a pit was dug for a water-wheel. Stiff bluish-grey Boulder Clay was exposed 
to a depth of 6 feet. Below the Boulder Clay shaly rock in situ was reached. The 
rock was much shattered just under the clay. Numerous boulders of the reddish- 
streaked rhyolite were picked out of the deposit. 

Near Llanor, 1\ miles inland from Pwllheli, and about 5 miles from Carnarvon 
Bay, there are brick- works in which the stiff bluish-grey Boulder Clay is exposed for 
a depth of about 20 feet. The bottom of the deposit is not reached in the section. The 
stiff clay is capped by 2 or 3 feet of yellowish weathered rubbly clay. Well-glaciated 
stones were common in the lower shift clay. The boulders consist chiefly of diabase, 
slate, etc., all of which appeared to be of Welsh origin. There were some indications 
of shells in an extremely fragmentary condition. The reddish-streaked rhyolite was 
seen imbedded in the clay. 

Passing west of a line from Pwllheli to Nevin the low grounds are underlaid by 
Boulder Clay and the smaller hills are covered by gravels and sands. Exposures are 
not common, but a few may be mentioned. On the hill immediately to the south of 
Tu-hivnt-i'r-afon Inn, Rhyd-y-clafdy, sand and gravel are exposed in many shallow- 
pits. Near the surface the material is gravelly, with a somewhat clayey matrix, and 
below this we find fine yellow marine-like sand. The exposures are only 4 to 5 feet 
deep, but the sands must attain a considerable thickness towards the base of the hill. 
Several gravel pits are found between Bodvean and Cors Geirch. One of these lies 
a quarter of a mile west of Bodvean, and shows a section 1 5 feet deep. It is situated 
on the slope of a hill at a height of almost 200 feet above sea-level. The deposits seen 
are mostly sands, fine, yellow in colour, and marine-like. Some pebbly layers occur, 
and at the top a little yellowish clay passing into soil. A little further south is found 
another with an exposure 6 to 7 feet deep. This is at a height of 175 feet above sea- 
level. It consists mostly of gravelly material getting finer and more sandy below. 
On the same hill slope, but at lower levels, fine yellow marine-like sands are exposed 
at several places. The entire hill is evidently covered by a mantle of gravel and sand. 






THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 43 

The marshy land below, known as Cors Geirch, is underlaid by Boulder Clay. This 
Cors probably marks the site of a former lake. 

At Nanhoron there is an old sand pit in the field in front of the house. It is now 
partly filled up, but formerly it was open to a depth of over 10 feet. Fine marine-like 
sands, well-bedded, are still exposed, with gritty layers here and there. These sands are 
of considerable thickness, for on digging to a depth of over 1 feet their base was not 
reached. The entire hill on which the house stands is rounded over by these sands. 
But to the south, where the Afon Horon runs at a lower level, the stream has exposed 
the lower stiff bluish-grey Boulder Clay which here, as in the coast sections, underlies 
the Sands and Gravels. The stream has here not cut through the Boulder Clay. 

Passing north to the neighbourhood of Madryn we still find similar features. In a 
field on rising ground some 200 yards south of the Hall, and 60 feet above it, sand and 
fine gravel are seen just below the surface. The sand is of the same character as that 
met with at Nanhoron. Again, in the park to the north of the Hall, fine yellow sand is 
exposed in many places and is often seen brought up from rabbit burrows. A large 
area is here covered by this sandy deposit. 

At Meillionen. near Ceidio, coarse gravelly sand is exposed at the surface and to 
a depth of 4 to 5 feet. In the lower ground adjoining, the stiff Boulder Clay is 
met with. 

At Cefn-leisog, in a field just south of the farmhouse, lies a disused sand pit. Here 
the sand is covered by 4 to 5 feet of yellowish weathered clay full of boulders. 

Passing to Brynodol we find that much of the land north of the house is sandy, as 
seen in several poor exposures and in newly dug ditches. 

At Ty-Issa, Tydweiliog, a boring was recently made for water. The deposits 
traversed were as follows : — 

3. Sand — 6 feet. 

2. Stiff bluish-grey Boulder Clay— 27 feet. 

1. Coarse sand — 3 feet. (The bottom of this sand was not reached.) 

A little higher up, at the County School, Tydweiliog, a boring was also made, and 
this passed through : — 

2. Sandy Clay— 16 feet. 

1. Pure Sand — 2 feet, the bottom of which was not reached. 
A little further down the slope of the same hill another boring was made which 
passed through nothing but marine-like sand to a depth of 10 feet. So we may gather 
from these various borings at Tydweiliog that the Drift covering the land in this 
neighbourhood consists of: — 

3. Sandy Clay. 
2. Sands. 

1. Stiff bluish-grey Boulder Clay, which at one place at any rate is underlaid by or 
includes Sands. 



44 DR T. J. JEHU ON 

Going westwards, at Plas-ym-mhenllech a sand pit was found in a field near the 
junction of the roads to Sarn and to Aberdaron. There were exposed here 5 feet of 
coarse sand in which were seen streaks of stiff clay. 

Going south again similar deposits are found at the surface in the neighbourhood of 
Sarn and Bottwnog. Near Llangian, just where the road turns off to Llawr-y-dref, 
there is a gravel pit in which 10 feet of coarse gravel are exposed. The gravels show 
faint indications of bedding. Another gravel pit is seen at Neigivl-Plas, in front of the 
house, the gravel being coarse and disclosed to a depth of 4 or 5 feet. All about the 
farm are many other disused pits of a similar nature. The gravels often pass down- 
wards into fine sands. A pit at Neigwl Ganol shows stony yellow weathered clay at 
the surface and to a depth of 4 feet. In a gravel pit at Talsarn the material is more 
rubbly — the stones being less rounded and more flaky. Above Talsarn Bridge the Afon 
Horon is seen to run on the stiff bluish-grey Lower Boulder Clay. This can be traced 
right up the stream, which further up its course becomes a mere rill and does not appear 
anywhere to reach the solid rock. 

The area between Abersoch and Porth Nigel is largely covered with sands. These 
are well exposed near Cim, at a height of over 200 feet above sea-level. The sands are 
fine and marine-like and have a yellow colour. 

Rhos-hirivawn, south-west of Llangwnadl, seems to be underlain by a stony 
weathered Boulder Clay. As we approach the extremity of Lleyn the mantle of Drift 
tends to become thinner until a mere skin of yellow weathered stony clay is left. 

No systematic examination was made of the erratics lying on the surface of the land, 
but in traversing the inland region in search of sections of the Drift the following were 
noted : — 

On Y Foel, between Llanllyfni and Clynnog Fawr, a block of felsite lies perched on 
the slate at a height of over 600 feet above the sea-level. 

On Y Gym Goch, at a height of 8()0 feet, a small boulder of the Eskdale granite 
was found, and a little lower down another of a granite from the south-west of 
Scotland. 

East of Porth Nant, on the western slope of Yr Eijl, a boulder of Carboniferous 
Limestone was picked up at a height of 450 feet. 

Near Tan-y-foel, between Llithfaen and Pistyll, a boulder of greyish-white granite, 
probably Scottish, was obtained at an elevation of over 500 feet. 

At the gate leading to Berth daur-fawr, near Penllech, there stands a huge boulder 
measuring 5 feet by 3 feet by 1^- feet. It is a very fresh-looking basic rock. A 
fragment was taken away and subsequently sliced and examined under the microscope. 
It proved to be a beautifully fresh olivine dolerite showing typical ophitic structure. 
Jt is difficult to locate its place of origin. The large size of the boulder makes one 
doubt whether it can have come from the west of Scotland, where similar rocks are found. 
It may possibly have been derived from some Tertiary dyke in Anglesey. 

On Penllech were found erratics consisting of granites, a quartz-porphyry with the 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 45 

quartz showing fine pyramidal forms, the reddish-streaked rhyolite, Millstone Grit, and 
quartzites probably from Anglesey. 

The reddish-streaked rhyolite again occurred on Mynydd Annelog, near the 
extremity of Lleyn, at a height of 350 feet, together with a boulder of Millstone Grit. 

Between Llanllawen and Aberdaron many boulders of diabase are dotted about. 
These are probably mostly of local origin. Blocks of jasper and of serpentinous 
rocks also occur which are derived from the neighbourhood of Careg, not far away 
to the north. 

On Mynydd Cilan, lying between Porth Nigel and Porth Ceiriad, and at heights 
of about 300 feet, the following erratics were found : — Several boulders of a coarsely 
crystalline diabase, the reddish-streaked rhyolite, Carboniferous Limestone with crinoid 
stems, a coarse grit quite unlike the rock of the district, light-coloured felspar porphyry 
derived from the masses of the rock which are found in situ to the north-east, Millstone 
Grit, and a reddish agglomerate. 

On Mynydd Mynytho, which consists of felspar porphyry, several erratics of diabase 
were found, also a boulder of a dark felsitic rock at a height of 400 feet on the south side 
of the mountain, and near it a small boulder of the reddish-streaked rhyolite. 

The reddish-streaked rhyolite is perhaps the commonest of all the boulders occurring 
in the Drift of Lleyn. It is found on the surface, in gravel pits inland, in the coast 
sections, and as pebbles on the modern beach. Its parent rock cannot be very far away, 
and probably lies to the north-east in Anglesey or Carnarvonshire. The boulders of 
this rock have been submitted to several petrologists, but no one has yet located its 
parent rock. 

IV. Summary and General Conclusions. 

In the western part of the Lleyn peninsula the Lower Boulder Clay is seen at some 
places to be separated from the bed-rock by an accumulation of angular detritus having 
all the characteristics of a " Eock Rubble " or " Head " (fig. 4). This rubble is disclosed 
in some of the coast sections towards the extremity of Lleyn, as, notably, at Porth Oer 
and at Aberdaron. It rests on the rocky slopes of the valley sides, where the valleys 
terminate seawards, and with these slopes it descends on either side below the level of 
the modern beach. This "Head" consists of angular blocks and fragments of all sizes 
packed close together or lying in an earthy matrix. The material seems to be entirely 
of local origin, and to be derived in the main from the immediately adjoining rocks 
above. It sometimes shows a rude bedding, as, for instance, at the east end of Aber- 
daron Bay, where it consists below of flakes and slivers of rock arranged with their long 
axes parallel to one another, giving a linear arrangement, and followed above by coarser 
material in which big angular blocks lie. During the time of the accumulation of this 
rock rubble the land must have stood higher relatively to the sea than it does at present. 
The lower ends of the valleys or inlets on the sides of which the rubble is seen are now 
often choked with the Drift deposits which pass below the level of the modern beach. 



46 DR T. J. JEHU ON 

The seaward terminations of these valleys or depressions are drowned, and their lower 
reaches must have been eroded during pre-Glacial times, when the land stood higher and 
extended further out to the sea than it does at present. The rook rubble is the result 
of subaerial work under severe climatic conditions. It is generally admitted that ice 
in some form or other has aided in the formation of such deposits. The rock fragments 
have all the appearance of being frost-riven, but apart from this there is not much trace 
of weathering. This " Head " may therefore be regarded as angular debris of local deriva- 
tion moved forward or downward by the action of frost, melting snows, and thawing sub- 
soils. On the steeper slopes the action of gravity alone would account for the slipping 
down of the material. Sometimes, on the other hand, the rubble is a confused mass of 
broken rock showing no arrangement. 

As a rule this rock rubble is overlain by the Lower Boulder Clay, but there are places 
where the Lower Boulder Clay seems to be absent and where the rubble is immediately 
followed by the LTpper Drift deposits. The rubble always lies directly on the bed-rock, 
and the fact that it is generally seen to have the Lower Boulder Clay superposed proves 
that it is the lowest of the superficial deposits which occur in the area. In the eastern 
part of the region this rock rubble is not seen unless it be represented by the shattered 
rock which underlies the Lower Boulder Clay at some places. 

The Lower Boulder Clay sometimes rests directly on the bed-rock, and where this is 
the case the rock surface as it passes under the clay has a markedly glaciated aspect ; some- 
times it is separated from the rock by a foot or so of broken and confused rock material, 
and sometimes, as on the rocky slopes at the seaward termination of valleys or inlets, it 
rests on an accumulation of rock rubble or " Head " of a considerable thickness. It is 
the most widespread of all the Drift deposits. This clay is very tough, compact, and 
homogeneous, having a massive appearance as displayed on the cliff face. It 
is often seen to be traversed by joints, and to break off in huge slabs along the joint 
faces as at Porth Nigel. The colour is dark bluish-grey, but after drying it becomes 
more of a light bluish-grey. Boulders are scattered sparingly through it, and are sub- 
angular or more or less rounded and often intensely glaciated. Distant erratics are 
common, and include Carboniferous Limestone, schists, etc., from Anglesey, various 
granites and porphyrites from the south of Scotland, and possibly some granites and 
lavas from the Lake District. It is interesting to note the occurrence of the Ailsa 
Craig riebeckite micro-granite at many places in this Boulder Clay and the frequent 
presence of chalk-flints which must have come from Cretaceous deposits on the floor 
of the Irish Sea or in the north-east of Ireland. The Boulder Clay is shelly, 
especially in the western portion of the peninsula ; and though the shell-remains are 
generally very fragmentary, some complete shells were obtained and a considerable 
number of forms have been identified. These include some species which are Arctic 
and Scandinavian and do not live now in British Seas, such as Trophon (Boreotrophon) 
clathratus, Trophon (Boreotrophon) scalariformis, and Astarte (Tridonta) arctica; 
some of a northern type but not confined to high latitudes, such as Cyprina islandica 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 47 

Astarte compressa, and Buccinum undatum ; and several species which are now 
common in British Seas. Thus we have in the Lower Boulder Clay a mixture of 
species belonging to different climates. These included shells and shell-fragments are 
therefore evidently derived, and point to the fact that the mer de glace of which this 
Boulder Clay is a product travelled over a sea-floor ere it reached this region. A part 
of a Belemnite was also found in the Boulder Clay at Aberdaron and several pieces of 
wood in that at Porth Nigel. 

This Boulder Clay is, as a rule, strikingly homogeneous and uniform in character 
and has evidently been subjected to great pressure. But at places a fine lamination is 
seen in the clay, especially towards the top. Where this lamination occurs the clay is 
stoneless and sometimes sandy. It appears to be sometimes due to shearing move- 
ments. Wherever land-sliding has taken place lamination occurs towards the top of 
the Lower Boulder Clay. On the foreshore at Porth Nigel an unctuous stoneless clay 
is seen which seems to result from the running or sliding downwards of Boulder Clay 
from which the coarser material has been washed out, leaving the residue in a buttery 
condition with a tendency to go on sliding down the foreshore. The lamination of the 
stoneless clay in the upper part of the cliff at Porth Nigel may have arisen in some- 
what the same way, for it is seen where slipping has taken place. 

The best exposures of the Lower Boulder Clay are met with in the coast sections, 
especially towards the western end of the Lleyn peninsula, but throughout the district 
there are indications that it underlies the Sands and Gravels, although these often form 
the lower part of the cliffs further east. The Lower Boulder Clay is shown by a boring 
to underlie the modern alluvium which forms the flats to the north of Dinas Dinlle. It 
has been detected below the Sands and Gravels of the Dinas Dinlle mound, but no 
further exposure is seen along this coast until we reach Gwydir Bay, where it 
forms a conspicuous feature below the other Drift deposits and under the foreshore. 
From Gwydir Bay westwards it appears in most of the bays, becoming more and more 
prominent and forming greater parts of the cliffs of Drift as we pass towards the 
extremity of Lleyn. It appears to have choked up most of the depressions, channels, 
and inlets met with along the coasts of the western part of Lleyn, sometimes even to 
the brim. These drift-filled depressions are now being rapidly eroded by the sea. At 
the head of the bays the Boulder Clay often forms the entire cliff, and extends below 
the modern beach to an unknown depth ; but towards the sides rock rubble is often 
seen lying between the Boulder Clay and the rocky slopes where these come down to 
the beach. At Porth Oer, Aberdaron, and Porth Nigel, cliffs formed of the Lower 
Boulder Clay attain heights approaching 100 feet. At Aberdaron, towards the west 
end of the bay, there appear to be inclusions of large irregular masses of sand in the clay, 
and towards the east end some bedded sands and gravel appear in the Lower Boulder 
Clay. The finest exposure of the Lower Boulder Clay in Western Carnarvonshire is 
that at Porth Nigel, where it forms high cliffs for a length of 3 to 4 miles. Following 
the coast on the southern side of the peninsula eastwards, the Lower Boulder Clay is 



48 DR T. J. JEHU ON 

last seen at Llanbedrog, where it forms the lower part of the cliff and the foreshore at 
the west end of the bay. A boring indicates its presence below the modern alluvium 
in the harbour at Pwllheli, but east of this it does not appear — the cliffs from Afon- 
wen to Criccieth being formed entirely of the Sands and Gravels and the Upper 
Boulder Clay. But though not exposed anywhere along this part of the coast, it 
probably underlies these Upper Drift deposits. 

Inland the Lower Boulder Clay is rarely exposed at the surface, but it may occa- 
sionally be seen underlying the marshy low grounds and moors or along the banks 
and beds of streams which have cut through the Upper Drift deposits, as, for instance, 
along the Afon Horon, below Nanhoron. Even where there are no exposures at the 
surface, borings and artificial cuttings prove it to be present under the other deposits. 
At Ty-Cam, Four Crosses, north of Pwllheli, it was found resting on a shattered surface 
of rock. Above the rocky cliffs at Porth Pistyll, south-west of Aberdaron, it is seen 
resting on the bed-rock at a height of over 100 feet above the sea-level. In other 
borings inland the bottom of this clay is not reached, but at Tydweiliog some sand was 
reached lying in or below the Boulder Clay. Shell-remains were not observed in the 
Boulder Clay away from the coast, except doubtfully in the brickfield at Llanor, north 
of Pwllheli. But the inland sections and exposures of this clay were too poor to render 
any adequate search for shell-remains or erratics possible. All that can be stated with 
certainty is that Boulder Clay underlies the sands and gravels in the interior as well as 
on the coast, and that this clay is of a dark bluish-grey colour and is very tenacious and 
compact ; that in some places it is full of boulders, many of which are well glaciated, 
and at others fairly free of boulders. The included boulders are most numerous in the 
eastern part of the district which approaches the mountains of Snowdonia, and these are 
mostly of Welsh origin. The Lower Boulder Clay has all the characteristics of a true 
bottom-moraine, and is undoubtedly the product of an ice-sheet which came from the 
north and invaded Western Carnarvonshire after having made its way over the bed of the 
Irish Sea and passed over Anglesey. The direction of transport of the included erratics, 
the configuration of the glaciated hills and the roches moutonnees, and the course of the 
rock-strise all indicate that the movement of the ice was roughly from north-east to 
south-west. Professor James Geikie, in his classic book — The Great Ice Age — gives a 
map of the British Isles during the Epoch of Maximum Glaciation in which he depicts 
this mer de glace as overwhelming Anglesea, skirting the north-west coast of Carnarvon- 
shire, and passing south-westwards over the tip of the Lleyn peninsula. The late Mr 
Carvill Lewis, on the other hand, in a sketch-map which is reproduced in Professor 
Bonney's Ice Work (1896), p. 18G, makes what he terms the Irish Sea Glacier terminate 
in an edge parsing along and just within the north-west boundary of Lleyn westwards 
towards Ireland, leaving the south of the Lleyn peninsula and Cardigan Bay outside its 
boundary. The facts recorded in this paper prove that this great ice-sheet covered the 
greater part of Western Carnarvonshire, passing over Lleyn into Cardigan Bay. 

In a previous paper communicated to this Society, the writer proved that this 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 49 

ice-sheet filled Cardigan Bay and even invaded Pembrokeshire. It is impossible to 
define accurately the eastern boundary of this ice-sheet. The northern ice met and 
overcame the Welsh glaciers which moved towards Anglesey from the valleys of 
Snowdon ; it invaded the foot-hills, such as Moel Tryfan, lying west of the Snowdonian 
mass. Professor Kendall, in Wright's Man and the Glacial Period (1893), p. 171, has 
pointed out that it was the proximity of Snowdon which really enabled foreign ice to 
invade Wales in the Moel Tryfan region. He observes that " a glance at the map will 
show that the ' radiant point ' of the Welsh ice was situated on or near Arenig Mawr, 
and that the great mass of Snowdon stands quite at the periphery of the mountainous 
regions of North Wales, so that it would oppose its bulk to fend off the native ice-sheet 
and prevent it from extending seaward in that direction." Consequently, the northern 
ice-sheet was able to overwhelm Western Carnarvonshire as far east as a line running from 
Carnarvon to Cardigan Bay somewhere in the neighbourhood of Pwllheli. The western 
slopes of the Snowdonian Mountains could only have given rise to small glaciers, and 
these would unite and become merged into the glacial mer de glace which moved south- 
westwards. South of Snowdonia a great glacier came down the Vale of Gwynant past 
Bedd-gelert, reaching Cardigan Bay at Traeth Mawr, where it was joined by another 
issuing from the mountains of Merioneth by the Vale of Ffestiniog. The united mass 
would be large enough to fend off the northern ice and to spread out as a fan towards 
Criccieth and Afonwen, and the volume of native ice would be increased by a smaller 
stream coming down the mountain valley between Moel Hebog and Carnedd Goch. 
The presence of the large volume of native ice in that neighbourhood explains the 
absence of foreign erratics and of shell-remains in the Drift deposits of the coast sections 
from Afonwen to Criccieth. The Sands and Gravels there seen are probably derived 
from a Lower Boulder Clay, although this is nowhere exposed on this part of the coast. 
We may assume that, if exposed, the Lower Boulder Clay in this part, like the Sands 
and Gravels, would show a general absence of foreign erratics. 

The Intermediate Gravels and Sands form a very variable series and are somewhat 
inconstant in their occurrence. They comprise both stratified and unstratified material, 
but the former predominates, especially in the coast sections. The best exposures are 
met with in the cliffs along the coast- line, the series being especially well developed at 
Dinas Dinlle, in the stretch between Clynnog and Gwydir Bay, at Nevin and Porth- 
dinlleyn, at the eastern end of Porth Nigel, at Llanbedrog, and east of Afonwen. The 
sands are, as a rule, yellowish or yellowish-brown in colour, and are marine-like in 
aspect ; they vary in texture, showing all grades from very fine sand to coarse gritty 
material. The coarser sands pass into gravel and boulder gravel. 

Mr Mellard Reade's investigations proved that the sand contains a large propor- 
tion of extremely well-rounded grains often possessing an extraordinary polish and 
transparency. He speaks of these as "microscopic boulders," and states that the 
contrast between these and the local grains is most remarkable. The sands and gravels 
usually show bedding, although at places confused irregular masses occur, especially of 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 2). 7 



50 DR T. J. JEHU ON 

the boulder gravel. As a rule the sands underlie the gravels. The pebbles and 
boulders of the gravels have not the flattened, ovoid form which is so characteristic of 
the shingle on modern beaches. At some places along the margin of the north-west 
coast, notably at Nevin, there occur laminated silty clays usually high up in the series. 
These silty clays are stoneless and devoid of shell-remains and appear to have been laid 
down tranquilly in water of some depth. Along this part of the coast the Drift consists 
chiefly of gravel and sand and silty clays, and the upper surface of the Drift has a 
terraced aspect. Some traces of disturbance and contortions are occasionally seen in 
the sandy beds. 

The sandy series at some places lies immediately on the bed-rock, but at others and 
especially in the deeper depressions, it rests on the Lower Boulder Clay. The line of 
demarcation between the Lower Clay and the Sands and Gravels is often well marked, 
but at places there appears to be a gradual passage from the one to the other. The 
pebbles and boulders in the Gravels include distant erratics of like origin to those 
found in the Lower Boulder Clay. Chalk-flints are not uncommon, and specimens of 
the Ailsa Craig riebeckite granite were obtained at several places. Shell-remains occur 
especially in the coarse sands and beds of gritty material. They are commonest in the 
sands on the north-west coast. Neither shell-remains nor distant erratics were noted in 
the Sands and Gravels on the south coast east of Llanbedrog. The shells found in the sands 
are very broken, usually occurring as mere crumbs. A few of those obtained from the 
sands at Nevin and Porth-dinlleyn have been identified, but the list is too limited to 
yield definite conclusions as to climate. But the sands in this region are very similar 
to the shell-bearing sands of Moel Try fan, and further south in Pembrokeshire,* the writer 
collected in similar sands a much larger number of shell-remains which were capable of 
identification. At both places the shells pointed to a mixture of species belonging to 
different climates. This, together with the fact that everywhere the shells are very 
broken and much rolled, indicates that they are not now in or near the positions in 
which they lived. In no case were Lamellibranch shells found with the two valves in 
apposition such as one finds in a recent and undisturbed alluvium like that which occurs 
to the north of Dinas Dinlle. Marine-like sands of a similar nature with gravels above 
are found inland in patches all over the peninsula. They lie on the slopes of, and 
sometimes spread over, the lower hills. Eastwards, as we approach the Snowdonian 
Mountains, the sands tend to become coarser and more greyish in colour and they are less 
marine-like. No shell-remains were noted in the sands of the interior, but the exposures 
met with are very poor and afford but little opportunity for a thorough search for shell- 
fragments. Such fragments are known to occur in the sands on Moel Tryfan and again 
away from the coast in Pembrokeshire, and so it is probable that this is the case also in Lleyn. 
Stratified sands and gravels were found at Llanaelhaiarn at a height of 450 feet above sea- 
level. The well-known shell-bearing sands of Moel Tryfan occur at a height of 1350 feet. 
In Pembrokeshire marine-like sands and gravels were found at elevations of 640 feet. 

* Trans. Roy. Soc. Edin., vol. xli., part i. (No. 4), 1904. 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 51 

These shelly sands and gravels evidently represent the materials of a sea-bottom 
carried onwards and upwards by an ice-sheet and rearranged by fluvio-glacial action. 
That is to say, they are the remanies of the great ice-sheet which came from the north 
and passed over the floor of the Irish Sea. The materials were laid down during the 
retreat of the ice-front, and at that time the lower parts of Lleyn must have been under 
water. The boulder gravel may be partly due to the action of torrential waters issuing 
from the melting ice-sheet and glaciers. No deposits indicating a genial inter-glacial 
epoch have been discovered in Western Carnarvonshire ; but this is not a matter for 
surprise, for such deposits, if they once existed, would stand very little chance of being 
preserved in a region so exposed as this. Whether the sands and gravels indicate a 
mere temporary retreat of the ice-sheet or the oncoming of a true inter-glacial epoch 
it is impossible to decide from the evidence afforded by a study of this region alone. 

The Upper Boulder Clay again is very sporadic in its occurrence and varies much 
in thickness and character. Sometimes it is a typical tumultuous unstratified till with 
boulders of all shapes and sizes scattered pell-mell throughout the clayey matrix ; at 
other places it has some of the character of rubble drift. As compared with the Lower 
Boulder Clay the Upper Till is less compact and homogeneous, more stony in character, 
and more weathered and friable. It often passes into a sandy clay. Where it becomes 
a rubble drift it consists of an agglomeration of subangular and more or less rounded 
boulders crowded thickly together in earthy debris. The included boulders are 
derived in the main from Welsh rocks, but many far-travelled stones are also found. 
Ice-scratched boulders are fairly common. The distant erratics are of similar origin to 
those found in the lower deposits, and include Scottish granites, the Ailsa Craig rock, 
and chalk-flints. 

Shell -fragments are again present, but these are scarcer and even less well preserved 
than those found in the Lower Boulder Clay. This Upper Boulder Clay is well 
displayed in the coast cliffs of Dinas Dinlle, and again between Point Maen Dulan and 
Clynnog, where it often forms the entire cliff. West of Clynnog it is not so conspicuous 
in the coast sections, but in many of the bays it can be seen to cap the cliffs which 
are formed of the Drift deposits. As a rule the Upper Clay rests on the Gravel 
and Sand series, but occasionally the two boulder clays occur together, the Upper 
Clay resting directly on the Lower, and in such cases it is sometimes difficult to 
separate them. 

In some of the smaller bays on the south-west coast of Lleyn the Lower Boulder 
Clay is not exposed ; in the cliff sections and at the sides of these bays the Upper 
Boulder Clay rests immediately on the rocky slope or is separated from it by some rock 
rubble. East of Afonwen the Upper Boulder Clay becomes again conspicuous. There 
it forms the upper part of the cliff, becoming more prominent and well-defined as we 
approach Criccieth. To the east of the town it forms the entire cliff. In the coast 
sections east of Afonwen no shell-remains were noted in this clay, and the included 
boulders seem to be entirely of native origin. 



52 DK T. J. JEHU ON 

Inland this Upper Boulder Clay is often absent. When it occurs it forms a thin 
covering only a few feet deep, and is often a mere rubble of boulders set in an earthy 
matrix. It has evidently suffered much from weathering and denudation. 

This Upper Boulder Clay may be the product of another mer de glace which 
followed the same course as that which produced the Lower Boulder Clay, and 
which was not much less extensive than its predecessor; or it may be due to the 
re-advance of one and the same mer de glace. Whether the two boulder-clays be the 
product of two separate and distinct ice-sheets separated by a genial inter-glacial 
epoch, or the product of one glacial epoch which was marked by an extensive 
oscillation of one ice-sheet, it is impossible to decide on the evidence presented in 
Western Carnarvonshire. The Drift deposits of this area can be accounted for on 
either of these theories. 

The ridges and mounds of gravel and coarse grey sand which run from north to 
south from the neighbourhood of Penygroes to that of Brynkir represent morainic 
material re-arranged and re-sorted by fluvio-glacial action and under the influence of the 
torrential waters which must have flowed over the surface at the time of the final 
disappearance of the ice. They occur near the eastward margin of the mer de glace 
which came from the north and passed over Lleyn. 

At the time of maximum glaciation the mountains of Lleyn must have been almost, 
if not entirely, buried under the ice-sheet. On Y Gyrn Goch, south-west of Clynnog, 
Drift was noted at a height of 1200 feet above sea-level, and a boulder of a foreign 
granite at 800 feet. The highest peak of Yr Eifl (The Rivals) reaches an elevation of 
1849 feet, and it is possible that the summit of this mountain stood as a " nunatak 
above the surface of the ice. The upper part of Yr Eifl has not a glaciated aspect, but 
of course this may be due to subsequent weathering. 

It is interesting to compare the sequence of glacial deposits in the Welsh area on the 
eastern side of Snowdonia with that met with in Lleyn. Professor Kendall # gives the 
succession on the coast east of Little Orme's Head as follows : — 

4. Boulder Clay with northern erratics and shells. 
3. Sands and Gravels with northern erratics and shells. 
2. Boulder Clay with northern erratics and shells. 
1. Boulder Clay with Welsh erratics and no shells. 

A similar succession is to be seen in the Vale of Clwyd. In these districts the Lowest 
Boulder Clay (I) is entirely the product of Welsh ice, which spread without hindrance 
seawards in the early stages of glaciation prior to the advance of the northern ice-sheet. 
In Lleyn this lowest Boulder Clay of the eastern area seems to be represented by the 
rock rubble or "Head" which underlies the Boulder Clay (2) laid down on the 
advance of the northern ice-sheet. So west of Snowdonia the sequence is as follows : — 

* Wkight's Man and the Glacial Period, 1893, p. 148. 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 53 

4. Boulder Clay with northern erratics and shells. 
3. Sands and Gravels with northern erratics and shells. 
2. Boulder Clay with northern erratics and shells. 
1. Rock rubble or "Head." 

That the lowest Boulder Clay which is found to the east of Snowdonia, and which is 
the product of native ice, does not occur also in Lleyn, is explained by the fact that there 
are no gathering grounds for the production of glaciers of any size in that area, but the 
severe climatic conditions would favour the accumulation of the rock rubble or " Head." 

The Drift which borders the coast of Lleyn has often a markedly terraced aspect. 
This appearance is well seen in the neighbourhood of Clynnog, and again between Yr 
Eifi (The Rivals) and Porth-dinlleyn. The upper surface of the Drift in these places 
looks remarkably even as seen from a distance, and lies at a height of 100 to 150 feet. 
A similar terraced aspect was noted at Llanbedrog, but here the surface is only about 
40 feet above sea-level. 

Where the Drift of the coast shows this more or less even surface it is found to consist 
for most part of the Sands and Gravels, and this may be taken to indicate a submergence 
of the land to an extent of 150 feet or more. But the character of the shell-remains and 
the form of the boulders in the boulder gravels militate against the theory that these 
sands and gravels are ordinary marine deposits ; and where the cliffs are rocky there are 
no indications that the land stood lower relative to the sea than it does at present. It 
is more probable that these stratified Drift deposits were accumulated between the margin 
of the retreating ice and the land area. The present coastal plain would at that time be 
under fresh water, due to melting of the ice and to the ponding of the land drainage. 

Along certain parts of the north-west coast of Lleyn, notably at Porth Colman, a rock- 
shelf or platform is seen at a height of 10 to 15 feet above the level of the present beach, 
which simulates an old shore platform such as that which has been described by Messrs 
Wright and Muff on the south coast of Ireland.* The surface of this platform is clear 
of Drift for a few yards inwards from its seaward edge ; and it slopes gently upwards until 
it disappears under the Drift — usually the Lower Boulder Clay (fig. 2). This platform 
cannot be traced back to any well-marked pre-glacial cliff, nor is it overlaid anywhere 
by any undoubted shore-deposits. At its inner end, where it passes under the Boulder 
Clay, the surface often has a glaciated aspect. In early Glacial times the shore-line stood 
further out to sea than it does at present, and the land descended in a gentle slope to 
sea-level. The surface underwent severe glaciation and was buried under Drift. Since 
that time there has been some subsidence, and the cliff has also been cut back by marine 
erosion and the foreshore lowered. Where the land slopes gently to form a low cliff, the 
Drift has been cleared away for some distance from the edge during stormy weather 
and high spring tides, when the waves would be dashed against the face of the cliff to 
considerable heights. 

* Scientific Proc. Roy. Dublin Soc, vol. x. (N.S.), part ii. (1904), p. 253. 



54 DR T. J. JEHU ON 

On the south coast of Lleyn there are indications that during some pre-glacial period, 
possibly the Pliocene, the bays extended further inland than they do at present. What 
appears to be an ancient line of cliffs passes inland from the sides of Porth Nigel, by 
Llanengan and Llangian, towards Bottwnog, and back along the eastern side of Mynydd- 
y-Rhiw (fig. 10). The base of the old cliff is hidden under Drift, but its course 
follows roughly the 100 feet contour line. The Drift deposits which fill much of this 
ancient bay are now being rapidly eroded again. The Drift is rather higher on the 
coast at Porth Nigel than it is for some way iidand, and so the drainage has been 
altered. The Afon Soch, which appears to have a natural outlet in this bay, is deflected 
so as to reach the sea at Abersoch (see Map). The course of an ancient pre-glacial 
coast-line can be traced again in the region east of Llanbedrog. 

At the time of the retreat of the ice-sheet glacial lakes must have been formed in Lleyn, 
and there are indications of overflow channels at many places, as at Trevor, in Nant Bodlas 
near Nanhorn, and along the valley of the Soch between Llangian and Abersoch. Some 
subsidence must have taken place during Glacial times, but there is no evidence to show 
that the land stood at a lower level relatively to the sea than it does at present. After 
the passing away of glacial conditions the land area appears to have had a greater extent 
than it does now. Peat, with the roots of trees in situ, are occasionally exposed at 
low tides in Cardigan Bay. The most recent movement would therefore appear to be 
one of subsidence, and this view is supported by Welsh traditions regarding the loss of 
a land area known as Cantref Givaelod which formerly extended over much of what 
is now Cardigan Bay. 

Some geologists hold that from the onset of the land-ice to the time of its final 
disappearance there was a period of continuous glaciation, during which the former sea- 
basins were never emptied of their ice-sheets. This view has recently been strongly 
advocated by Mr Lamplugh. To account for the accumulation of the stratified deposits 
associated with the boulder clays in the West British area, somewhat extensive oscilla- 
tions of the mer de glace are admitted. Speaking of the Isle of Man in his Presidential 
Address to the Geological Section of the British Association (1906), he says that at the 
time of its maximum glaciation the surface of this ice-sheet stood more than 2000 feet 
higher than present sea-level. " Then followed a declining stage in glaciation, during 
which the ice-sheet shrank away from the hills, which were never again covered. . . . 
The recession of its margin was relatively accelerated in the northern part of the island, 
so that a broad hollow was formed there between the hills and the ice-border : in this 
hollow a mass of stratified drift was deposited." This stratified material he regards as 
marginal " overwash " from the ice-front. Though waning and shrinking for a time, he 
thinks the ice-sheet did not disappear entirely from the surrounding basin. " Following 
closely upon this local deposition of stratified drift, there appears to have been a limited 
re-advance of the ice, which brought about the accumulation of an upper boulder clay 
on parts of the low ground." He thus admits a declining stage of glaciation. which was 
followed after the deposition of the stratified material in a re-advance of the ice, but 



Trans. Roy. Soc. Edin. Vol. XLVIL 

L)r T. J. Jehi on "The Glacial Deposits of Western Carnarvonshire." 




To face p. 54. 



THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 55 

only to an extent sufficient to " close in again upon the lower flanks of the hills." In 
his Survey Memoir, on " The Geology of the Isle of Man " (1903), p. 395, Mr Lamplugh 
has adduced reasons to show that " the shrinkage of the ice-sheet covering the Isle of 
Man is likely to have commenced while the Welsh and Ivernian sheets were still in- 
creasing." "But," he adds, "it was not until there had been a great amelioration of 
climate that the island began to emerge from the waning mass " (p. 395). So this 
"great amelioration of climate" must have taken place at the time in which the 
stratified Drift deposits were accumulating between the receding ice-sheet and the 
emerging land. Does not this change in the climate indicate the gradual passing away 
of glacial conditions at a time prior to the deposition of the Upper Boulder Clay ? Mr 
Lamplugh refers again and again to the fact that a very considerable alteration in the 
climate must have occurred to account for the shrinkage of the ice from the island. 
" The shrinkage," he observes (p. 396), "may have begun from a diminution in the 
amount of snowfall alone ; but before it could proceed far there must have been a wide- 
spread change of climate and an essential difference between the conditions of this stage 
and of the initial stage of the period. At the beginning of the period the climate was 
such as to permit the permanent snow-line to descend nearly to sea-level ; while, when 
the ice-sheet had reached its maximum, it was requisite that the climate should be warm 
enough to prevent snow lying permanently at over 2000 feet above sea-level before a 
positive lowering of the surface by melting could take place. Thus, an ice-sheet 
already in existence may be able for a long time to withstand a climate in which it 
could never originally have accumulated ; and ameliorating changes may attain an 
advanced state before their effect on such an ice-sheet becomes marked. But when the 
permanent thawing of the surface once commences, it must go on with accelerated 
rapidity as lower levels are reached." Again, a little further on he says : " It is clear 
that by the time the hill-tops had reappeared the Arctic conditions of climate had 
passed away, and permanent snow was no longer possible on rock-surfaces, even at the 
higher levels." If the Arctic conditions of climate had thus passed away by the time 
the hill-tops had reappeared, it is likely that the amelioration in the climate continued, 
and that something more than a mere temporary and short-lived recession of the ice-lobe 
took place. Mr Lamplugh accounts for the deposition of the Upper Boulder Clay by 
a re-advance of the ice which followed closely again the local depositions of the stratified 
drift. In view of the great change of climate which is admitted, it seems more probable 
that at this time also the ice-plateau, though lingering for a considerable time in the 
Irish Sea basin, went on wasting and shrinking, passing into a condition of " dead ice," 
and eventually disappearing entirely in much the same way as Mr Lamplugh thinks the 
final waning of ice-sheets from the British area to have taken place. 

Though the mass of ice may not have attained its maximum development at the 
same time in all parts of the West British region, and shrinkage may have started in the 
ice covering the Isle of Man sooner than it did further south and west, one cannot 



56 DR T. J. JEHU ON THE GLACIAL DEPOSITS OF WESTERN CARNARVONSHIRE. 

imagine any recession of ice from the land area which was not practically contempor- 
aneous on all sides of the Irish Sea basin. 

When the ice-sheet which deposited the Lower Boulder Clay shrank away from the 
hills and the area now known as Lleyn emerged from its icy covering, the ice probably 
still lingered for a time in the basin of the Irish Sea. It was probably during this time 
of shrinkage and recession that the sands and gravels lying between the two boulder 
clays were accumulated. Owing to melting at the edge of the ice-sheet and to the 
ponding up of the land drainage, much of Western Carnarvonshire must for a time have 
been submerged under fresh water, and there is no need to assume any sinking of the 
land relatively to the sea of which there is no clear evidence in this region. The facts 
which militate against the supposition that these shelly sands and gravels are ordinary 
marine deposits laid down when the land stood lower relatively to the sea are summarised 
by Professor Kendall in Caevill Lewis' Glacial Geology of Great Britain and 
Ireland (1894), Appendix viii. 

Stratified Drift deposits — shell-bearing sands and gravels — are met with lying 
between two boulder clays at various elevations on both sides of the Irish Sea — in 
Ireland as well as in West Lancashire, Cheshire and Wales. If these deposits were laid 
down in their present position during a retreat of the ice-sheet from the hills, it means 
that a shrinkage took place from the land area on both sides of the Irish Sea as well as 
about the Isle of Man. The recession of the ice would be accelerated by the action of 
land streams impinging on to ice, and wasting at the surface would become more rapid 
as the surface was brought lower. The shrinking mass would thus be attacked on all 
sides. Granting, therefore, as Mr Lamplugh does, a widespread change of climate to 
account for the shrinking of the ice-sheet from the Isle of Man, and recognising that a 
similar shrinkage must have taken place on all sides of the Irish Sea basin, the proba- 
bility is that the ice- plateau vanished entirely and that an inter-glacial epoch intervened 
between the times of the deposition of the two boulder clays. 

The Upper Boulder Clay would thus denote a recurrence of severe glacial conditions 
and the advance of another ice-sheet. It has already been stated that, as far as Lleyn 
is concerned, the two boulder clays might be regarded as the product of one mer de 
glace subject to considerable oscillation, but a review of the whole of the Irish Sea area 
renders it more probable that they are the products of the ice-sheets of two glacial 
epochs separated by an inter-glacial epoch. 

(The author is indebted to the Carnegie Trust for a Research Grant towards the 
expenses connected with this work.) 



Trans. Roy. Soc. Edin. 



Vol. XLVII. 



Dr T. J. Jehu on " The Glacial Deposits of Western Carnarvonshire." 




w 
o 



< 
o 

to 



o 




1- 


h- 


z 




LU 


Ui 


< 




Ul 


u 

U. 






O 


o 


u_ 




o 


o 


o 




f 


OJ 


1— 
X 
03 




at:/.-: 














- 






X 






i 



( 57 ) 



III. — Scottish National Antarctic Expedition : Observations on the Anatomy of the 
Weddell Seal (Leptonychotes Weddelli). By David Hepburn, M.D., Professor of 
Anatomy, University College, Cardiff (University of Wales). (With One Plate.) 

(MS. received February 22, 1909. Issued separately May 25, 1909.) 



Introductory. 

The anatomy of all marine mammals presents an interesting field of observation 
upon their structural adaptation to a particular environment, and naturally, therefore, 
an extensive literature already exists in respect to these mammals, notwithstanding the 
many difficulties connected with theif detailed examination. Their large size and the 
rapidity with which their tissues undergo decomposition have been serious obstacles in 
the way of prolonged dissection both of Cetacea and of Seals. Consequently, the 
examination of many parts of their anatomy has, of necessity, been more or less hurried. 
Accordingly, in the course of the preparations for the Scottish Antarctic Expedition, 
led by W. S. Bruce, LL.D., etc., arrangements were made for the preservation of 
mammalian specimens by the same injection methods that are now commonly used for 
the practical study of human anatomy, and the medical officer to the Expedition 
received practical instruction from myself in the application of these methods. 

The animal now under consideration was preserved in the following manner : — One 
of the vessels between two of the toes on a hind flipper was opened and a fine canula 
inserted. Through this, a quantity of an arsenical preservative, containing glycerine, 
was introduced under gravitation, and to-day the tissues are as soft and free from 
putrefaction as they were on the day the creature was killed. 

The animal is a young male Weddell seal (Leptonychotes Weddelli), one of the 
earless seals. It was born on or before the 29th of August 1903, and was killed by 
hydrocyanic acid on the 31st of August 1903, and immediately injected with the pre- 
servative solution. 

Its dimensions are as follows : — 

From tip of snout to end of tail . 

,, ,, ,, hind nipper 

; , „ ,, fore-limb 

,, ,, genital aperture 

From genital aperture to anus 

,, anus to tip of tail . 
Interorbital width 
Axillary and greatest girth 
Fore flipper : — 

Length on outer edge 



„ „ inner „ . 
Stretch (expanded) . 
TRANS. ROY. SOC. EDIN., VOL. XL VII. PART I. 



(NO. 3) 



51* 


ins 


60£ 


J! 


13" 


IJ 


39 


)> 


8 


>) 


** 


J) 


3f 


!) 


34* 


)> 


11* 


>> 


H 


)) 



74 



58 PROFESSOR DAVID HEPBURN ON 

Hind flipper : — 

Length on outer edge . . . . . 13f ins. 

„ „ inner „ . . . . . . 12 ,, 

Stretch (expanded) . . . . . . 13J ,, 

Circumference at hase . . . . . 10 V ,, 

Its weight was 83 lbs. A portion of umbilical cord about 3 inches in length was 
still dangling from the umbilicus. Its fur was greyish yellow, mottled by darker spots, 
suggesting leopard characters. 

Abdominal Viscera, — On opening the abdomen, the general shape of the cavity was 
seen to be ovoid, being wider in front and narrowing behind to the well-marked inlet 
of the pelvic cavity. Regarded as a whole, the abdominal cavity presented a close 
resemblance to that of the porpoise,* of which an account appeared in the Transactions 
of the Royal Society of Edinburgh, vol. xl., part ii., p. 315. 

Running forwards from the umbilicus, the umbilical vein, already nearly obliterated, 
formed a very distinct object, suspended as it was in the free edge of an extensive 
ventral mesentery constituting the falciform ligament of the liver. Running back- 
wards from the umbilicus, the urinary bladder, with its associated hypogastric arteries, 
was likewise suspended in a mesial ventral mesentery whose depth from the abdominal 
wall to the bladder measured from an inch to an inch and a half. The bladder was 
long and narrow, and at the umbilicus its lumen was distinct. No part of the bladder 
had become obliterated to form the urachus, so that the bladder represented the entire 
intra-abdominal extent of the allantois. 

The liver and stomach occupied the wide anterior end of the abdominal cavity, and 
in rear of them only the mass of coiled intestine was visible. 

Peritoneal Folds and Reflections. 

At first glance there was no sign of a great omentum, but on separating the coils of 
intestine from contact with the stomach, the great omentum was found crumpled up 
against its attachment to the greater curve of the stomach. Its greatest depth from the 
stomach to its free margin was 3 inches. It was devoid of visible fat, and was there- 
fore an extremely thin and translucent membrane. There was a well-marked gastro- 
hepatic or lesser omentum presenting the usual gastric and hepatic connections. The 
gastro-splenic omentum was also well marked, and by it the spleen was attached to the 
great curve of the stomach. From the dorsal aspect of the spleen the peritoneal 
membrane extended to the dorsal wall of the abdominal cavity without coming into 
relation with the left kidney, which was situated opposite to its fellow and considerably 
further back in the abdomen. 

The liver was attached to the sub-diaphragmatic surface by the usual suspensory, 
coronary and lateral peritoneal ligaments, to which further reference will be made in 
describing the liver. 

* Hepburn and Waterston, loc. cit. 



OBSERVATIONS ON THE ANATOMY OF THE WEDDELL SEAL. 59 

The duodenum was suspended in a dorsal mesial mesentery whose base of attachment 
measured nearly 5 inches, and this represented the distance between the pylorus 
(gastro-duodenal junction) and the duodeno-jejunal junction which was situated close 
behind (tail wards of) the superior mesenteric vessels and immediately to the left of the 
mesial plane. To the right side of the duodenal mesentery there was a large peritoneal 
recess whose right boundary was formed by a tailed lobe of the liver extending back- 
wards along the dorsal wall in intimate association with the inferior vena cava. 

At the first glance there appeared to be no naked-eye distinction between the small 
intestine and the colon. At no point were taenia coli, sacculations, or appendices 
epiploicse visible. 

As measured from the duodeno-jejunal junction or flexure, 46 feet of gut were 
suspended from the dorsal wall in a mesentery practically corresponding to the entrance 
of the superior mesenteric vessels. A closer examination revealed a distinct lateral 
diverticulum rather more than 1 inch in length but of the same calibre as the gut, 
situated on that part of the gut, which was suspended in the mesentery and at a 
point 9 inches from the hinder end of the mesentery. This diverticulum evidently 
represented the ccecum and the vermiform appendix in their most primitive form, and 
in the same condition as I have formerly described in the case of the grey seal 
{Halichcerus grypus).* We may therefore conclude that this diverticulum marks 
the commencement of the colon. 

Consequently the first 9 inches of the colon are suspended in the same mesentery 
as the small intestine, exclusive of the duodenum. The remainder of the gut, i.e. colon, 
was suspended in a dorsal mesial mesentery extending tailwards to the entrance of the 
pelvis, while through the greater part of the pelvic cavity a dorsal mesial mesentery 
supported the pelvic part of the colon, i.e. the rectum. 

The foramen of Winslow was very distinctly denned in relation to the free right 
border of the gastro-hepatic omentum. Similarly the lesser sac of the peritoneum was 
equally distinct, although the great omentum was not in any sense a gastro-colic structure. 

The inlet of the pelvis measured 1 inch in the transverse diameter and 3 inches 
in the conjugate diameter. Its lateral boundaries were well defined by the hypogastric 
(umbilical) arteries, each of which was supported in a peritoneal sling or ligament attached 
to the dorsal wall along the pelvic inlet. 

The pelvic peritoneal pouch extended backwards between the gut and the bladder to 
a distance of 3^ inches from the pubic crest, and so reached a point posterior to the 
bulb of the urethra, i.e. to a point which corresponded with the central point of the 
perineum. On the ventral abdominal wall, 1 inch to the outer side of the hypogastric 
arteries, there was the opening which led into the inguinal canal (see fig.). Each 
opening was circular and half an inch in diameter. It led into a circular peritoneal tube 
which extended through the abdominal wall in relation to the hinder free border of the 
musculus transversalis abdominis and the musculus obliquus interims abdominis. At 

* Hepburn, " The Grey Seal {Halichcerus grypus)" Jour. Anat. and Phys., vol. xxx. 



60 PROFESSOR DAVID HEPBURN ON 

the plane of the musculus obliquus externus abdominis, and close to its pubic attach- 
ment, the peritoneal tube passed through a muscular slit corresponding to the external 
inguinal ring and thereafter terminated in a blind end situated close to the posterior 
end of the testis. The length of this tube of peritoneum was 2^ inches, and its 
testicular part formed the tunica vaginalis testis. 

Each vas deferens entered the abdomen through the internal inguinal ring, being 
suspended from the ventral wall in a short mesentery or peritoneal sling by which each 
vas was carried across the abdominal aspect of the hypogastric artery to the pelvic 
surface of the urinary bladder about the level of the pubic crest. These two mesenteries 
for the vasa deferentia formed a free transverse fold of peritoneum on the pelvic surface 
of the urinary bladder at the level indicated. From this level the vasa deferentia passed 
backwards on the pelvic surface of the bladder under cover of the peritoneum. On the 
pubic aspect of the bladder there was a triangular non-peritoneal surface extending for- 
wards from the pubic crest for If inches, at which point the ventral mesentery of the 
bladder commenced and continued to the umbilicus. 

Alimentary Organs. 

The stomach (see h'g.) presented a single chamber situated with its long diameter 
in the axis of the trunk. The oesophagus entered the stomach slightly to the right side 
of the most prominent or anterior part of the fundus. The greater convexity or curve 
was well defined, and measured 17 inches from the oesophagus to the pylorus.- The 
distance between the same points along the lesser curve was 10 inches. Each of these 
curves was associated with the usual omenta. A very sharp bend occurred in the lesser 
curve, 5 inches in a straight line from the oesophagus and 3 inches from the pylorus. 
The sides of this bend were held in close apposition by the peritoneum, and the general 
appearance produced was that of a constriction in the course of the cavity of the 
stomach. The greatest width of the stomach on the oesophageal side of this bend was 
5^ inches, whereas on the pyloric side of the bend the greatest width was 4 inches, and 
opposite the bend the width was 2^ inches. The practical result of the infolding of the 
stomach wall was therefore to produce two chambers communicating with each other 
by an aperture considerably narrower than either of the chambers. A second slight 
constriction was present in the pyloric section, and thus as a whole, from oesophagus to 
pylorus, the stomach suggested three imperfectly separated compartments or chambers. 

The contents of the stomach consisted of a quantity of a thick pasty substance of a 
somewhat light earthy colour. It was uniformly smooth, and contained no evidence of 
bones. In all probability it represented partially digested coagulum of milk. 

The pylorus was placed in the mesial plane, and was recognisable both to the eye 
and to the touch as a constricted ring:. 

The duodenum extended from the pylorus in the form of a horse-shoe loop 12 to 
13 inches long. It was attached dorsally in the mesial plane by a mesentery. The 



OBSERVATIONS ON THE ANATOMY OF THE WEDDELL SEAL. 61 

proximal and distal ends of the loop were from 4 to 5 inches apart. To the right side 
of the duodenal mesentery there was a large peritoneal recess or pouch the mouth of 
which was directed anteriorly, and to which reference has already been made. 

The duodeno-jejunal flexure lay immediately to the left side of the vertebral column 
and marked the beginning of the mesentery proper and of the small intestine suspended 
in it. This mesentery had a very short base which practically corresponded to the 
entrance of the superior mesenteric vessels. It was twisted to the right and supported 
the jejunum and ileum, which together measured almost 46 feet in length. 

The coecal diverticulum formerly referred to as representing the combined coecum 
and vermiform appendix, was taken to indicate the termination of the small and the 
beginning of the large intestine, This primitive developmental form of the coecum 
and vermiform was rather more than an inch in length, while in calibre it corresponded 
with the gut. From this diverticulum to the end of the gut there were no other 
external evidences of any distinction between small and large intestine. 

From the coecum the large intestine or colon pursued the first 9 inches of its 
course suspended in the same mesentery as the small intestine. Thereafter the colon 
assumed a mesial position and, as far as the pelvic inlet, i.e. for a distance of 18 inches, 
it was suspended in a dorsal mesial mesentery. The pelvic portion of the colon was 
also placed mesially, and the greater part of it was also suspended in a dorsal mesentery. 
From the pelvic inlet to the anal aperture the gut measured nearly 1 inches, so that 
the entire length of the colon from coecum to anus was practically 3 feet. Thus the 
total length of the gut from pylorus to anus was : — 

Duodenum ...... 1 foot. 

Small intestine . . . . . .46 feet. 

Colon . . . . . . 3 ,, 

Total . . . . . 50 „ 

The liver (see figs.) was large and deeply fissured, thereby presenting very dis- 
tinct lobes. It was intimately associated with the diaphragm, and occupied the anterior 
end of the abdominal cavity from side to side. It was provided with the usual peritoneal 
ligaments. The falciform and coronary ligaments presented no special features as 
regards their arrangement, but the left lateral ligament extended from the sharp left 
margin of the liver whereas the right lateral ligament was short and extended from the 
smooth surface of the right aspect of the right lobe. The diaphragmatic surface of the 
liver was smooth and convex, adapting itself to the abdominal surface of the diaphragm 
and presenting right and left lobes in relation to the suspensory or falciform ligament. 
The right lobe was divided into mesial and lateral portions by a deep dorso-ventral 
fissure, and the left lobe was imperfectly divided by dorsal and ventral notches which, 
however, did not meet each other. 

On its visceral aspect the liver was much subdivided, particularly in relation to the 
right lobe. The right and left lobes were marked off from each other by the ligamentum 
teres (obliterated umbilical vein) on the ventral aspect of the hilum, and by the hepatic 



62 PROFESSOR DAVID HEPBURN ON 

sinus on the dorsal aspect of the hilum. The right lobe presented the same mesial and 
lateral portions which were noted on the diaphragmatic aspect. The mesial portion 
was divided into dorsal and ventral parts by the hilum. To the ventral portion the 
gall-bladder was attached, and this part of the right lobe was connected to the left lobe 
by a pons hepaticse. The dorsal portion was to a large extent concealed by the large 
hepatic sinus. The lateral part of the right lobe was also divided into dorsal and 
ventral segments by the extension of the hilum across its visceral aspect. 

A tailed lobe extending from the dorsal segment of the mesial portion of the right 
Lobe passed backwards on the ventral surface of the inferior vena cava, which was thus 
embedded in the liver substance. This tailed lobe expanded so as to conceal a consider- 
able length of the inferior vena cava, and into this hidden part of the cava there opened 
directly the hepatic veins from this particular lobe, as also the hepatic veins from the 
lateral part of the right lobe. 

The inferior vena cava itself opened into the large hepatic sinus situated close to the 
diaphragm and extending right and left of the suspensory ligament. This sinus 
received the hepatic veins from the right and left lobes of the liver, with the exception 
of those veins already mentioned as opening directly into the inferior vena cava. The 
interior of this sinus was imperfectly divided into right and left parts by a crescentic 
partition which was situated to the right side of the line of attachment of the suspensory 
ligament, so that the part of the sinus to the left side of the crescentic fold was the 
larger. The inferior vena cava opened into the hepatic sinus on the right side of the 
imperfect partition, while on its left side it received the mouth of the ductus venosus. 

The hepatic sinus narrowed for the purpose of passing through the diaphragm in 
order to enter the right auricle of the heart, but the general size of its lumen was so 
much greater than that of the inferior vena cava that it is more accurate to say that 
the inferior vena cava opened into the sinus and the sinus joined the auricle. 

The conditions presented by the hepatic veins afford some interesting light upon 
the question of their development. Clearly the large hepatic sinus has resulted from 
the fusion of the two embryonic venae revehentes, although, from the position of the 
imperfect crescentic partition, it is evident that the left vena revehens was the larger of 
the two and therefore received the smaller or right vena revehens. This arrangement 
would therefore appear to provide a variation upon the current statement that " the 
left vena revehens loses its connexion with the sinus venosus and opens into the right 
vena revehens" (Cunninghams Text-Book of Anatomy, 2nd ed., p. 935). Again, the 
ductus venosus is described as passing directly " from the left vena advehens to the right 
hepatic vein" (vide ibid.), whereas, in the liver under consideration, the mouth of the 
ductus venosus opens to the left side of the crescentic fold, which appears to represent 
the remains of a fusion between the right and left hepatic veins. Further, to quote 
again from the same authority, " The upper part of the inferior vena cava is developed 
as an outgrowth from the common trunk formed by the fusion of the ductus venosus 
with the right hepatic vein." From the present dissection the inferior vena cava would 



OBSERVATIONS ON THE ANATOMY OF THE WEDDELL SEAL. 63 

appear to arise from the right hepatic vein independently of the ductus venosus, more 
especially in view of the fact that the inferior vena cava receives direct tributaries from 
the caudate lobe adherent to its ventral aspect. 

The gall-bladder, which was empty, occupied a fossa on the visceral aspect of the 
mesial portion of the right lobe of the liver. It thus lay to the right side of the 
ligamentum teres, from which it was separated by a projecting portion of liver 
substance, connected to the left lobe by a pons hepaticse. This portion of the liver 
corresponded in general position with the quadrate lobe of the human liver. 

The cystic duct passed towards the hilum of the liver and, having been joined by 
hepatic ducts corresponding in number with the three chief lobes of the liver, the 
common bile-duct was formed. This duct pursued its course on the ventral aspect of 
the foramen of Winslow and, immediately after passing this foramen, i.e. just before 
reaching the duodenum, it was joined by the pancreatic duct. The duct thus resulting 
came in contact with the duodenum a little more than half an inch from the pylorus. 
It perforated the duodenal wall very obliquely, and opened on the summit of a papilla at 
a distance of 2 inches from the pylorus. 

The pancreas presented a characteristic appearance. A small part of this gland 
was found within the duodenal mesentery, but the greater part of the organ extended 
headwards behind the lesser peritoneal sac. The pancreatic duct emerged from the 
substance of the organ on the cephalic (anterior) side of the foramen of Winslow (not 
from that part within the duodenal mesentery), and, extending to the right side, it 
formed a junction with the common bile-duct on the cephalic side of the foramen of 
Winslow and fully 2 inches from the common orifice by which both ducts pour their 
contents into the duodenum. 

The spleen measured from 9 to 10 inches in length, from 1 to 6 inches in width, 
being narrow anteriorly and wide posteriorly. It was extremely thin, being not 
more than a quarter of an inch in thickness. It was situated between the stomach and 
the dorsal wall, being attached to the former by the gastro-splenic omentum and to the 
latter by a dorsal mesial mesentery. There was no intimate relation between the spleen 
and either of the kidneys, because these organs were removed from all immediate 
association by reason of their situation considerably nearer to the pelvic region. The 
tail end of the pancreas extended into the posterior splenic mesentery but did not come 
in contact with the spleen. 



Trans. Roy. Soc. Edin. Vol. XLVII. 

Prof. David Hepburn on " Observations on the Anatomy of the Weddell Seal." 



DORSAL. 



CESOPH. 



- 



Stomach 




LEFT. 





LIG. TERES. 

VENTRAL. 

Liver of seal. View from abdominal aspect. 



LOBE ON VENTRAL 
ASPECT OF INF. V. 
CAVA (TURNED TO- 
WARDS DORSUM). 



RIGHT. 




URINARY BLADDER. 



HYPOGASTRIC A. 



INT. 1NG. RING. 



VAS DEF. 



LIG. TERES. 

Liver of seal (seen from ventral aspect). 



Pelvic cavity of seal (seen from pelvic inlet). 



( 65 ) 



IV. — Supplementary Report on the Hydroids of the Scottish National Antarctic 
Expedition. By James Ritchie, M.A., B.Sc, Natural History Department, The 
Royal Scottish Museum. Communicated by W. S. Bruce, LL.D. 

(MS. received December 8, 1908. Read January 4, 1909. Issued separately May 27, 1909.) 

Since the report on the Hydroids collected by the Scotia was completed, almost 
three years ago, I have had an opportunity, thanks to the kindness of Dr W. S. Bruce, 
of looking through the enormous mass of material brought together by him during his 
Antarctic voyages. The search, begun in the hope of finding a few minute species— 
which, since many are epizoic on other forms of marine life, might readily be overlooked 
on cursory examination — has resulted in the addition of no less than 25 forms to the 
36 already recorded. Deducting from this total of 61 forms two varieties, it appears 
that the Scotia collection contains in all 59 distinct species. Few, indeed, of these have 
been obtained in truly Antarctic latitudes — subantarctic or temperate seas having 
furnished the greater number — but this result is in accordance with the findings of the 
majority of the recent Antarctic expeditions.* Dr Billard, in summarising the results 
of the Belgian, Swedish, and French expeditions, states that the known Hydroid fauna 
of Antarctic regions comprises only 32 species. But from his list he has, I think un- 
justly, excluded South Georgian records. His list, moreover, published before the 
valuable paper by Professor Hickson and Mr Gravely on the Discovery Hydroids was 
issued, necessarily makes no reference to the large number of forms recorded by them 
from Victoria Land. 

The following figures indicate the wealth in species of the Hydroid fauna of Antarctic 
seas, the Antarctic area being defined as the region lying in "higher southern latitudes 
than the extreme limit of floating ice, according to the most recent research " (Bruce, 
1894, p. 208); that is, an area corresponding to the Antarctic Circumpolar Subregion 
of Ortmann (1906). To this region belong the 14 species collected by the Belgian 
expedition (Hartlaub, 1904), 24 of the Swedish collection of 50 species (Jaderholm, 
1905), the 8 species of the French expedition (Billard, 1906, (2) ), and 24 of the 25 
English records (Hickson and Gravely, 1907). From the same area the Scotia 
obtained only 9 species, although some of the most fruitful stations, Burdwood Bank, 
the Falkland Islands, and Go ugh Island, lay just beyond its boundary. Extracting 
the distinct species from those lists, we find that their number amounts to 68. 

The Hydroid Zoophytes, then, are represented in the Antarctic circumpolar seas, 

* The report on the Hydroids of the German expedition has not yet been published. 
TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 4). 9 



66 MR JAMES RITCHIE : SUPPLEMENTARY REPORT ON 

according to our present knowledge, by 68 different species, 21 of which are Gymnoblasts, 
47 Calyptoblasts, while of the total, 36 must, so far, be regarded as peculiar to the 
Antarctic region. As regards the number of individuals as distinct from variety of 
species, recent results indicate that the area is a thinly populated one, for comparatively 
few Hydroids were brought up in any one haul of trawl or dredge, a strong contrast' to 
the abundance of tangled masses which occur in such temperate areas as the North Sea. 
On the whole, the Antarctic Hydroid fauna may be looked upon as a rather meagre 
one, comprising comparatively few species belonging to well-established genera. 

From the point of view of geographical distribution the collection to be described 
is of a miscellaneous character, for Dr Bruce collected material on his outward 
and homeward voyages as well as at his southern stations. Thus specimens from 
tropical seas — from the Cape Verde Islands, from Brazil, and from St Helena — are 
recorded alongside specimens from subantarctic and Antarctic localities, but so little 
is known regarding the Hydroid fauna of those places that it seemed better to include 
them here, at the same time furnishing a record of all the Hydroids brought together 
by the Scotia. 

Two species, Thyroscyphus tridentatus and Plumularia lagenifera, hitherto recorded 
only from the South and North Pacific respectively, have been found in the South 
Atlantic, while the known ranges of several other species have been considerably 
extended. 

Of new forms there have been described the trophosome and gonangium of 
Sertularia heterodonta from off Brazil, the trophosomes of Antenella quadriaurita 
from Gough Island, and of a variety of Lafoea gracillima from deep water to the 
south of the South Orkneys, while the gonangia of Sertularia rathbuni and Anten- 
nopsis scotix have been described for the first time. 

Some additions have been made to the existing descriptions of rare species, and 
noteworthy variations have been recorded in the hope that the realisation of the con- 
siderable limits within which variation is possible may tend to the better understanding 
of reliable specific characters. 

I gladly seize this opportunity of setting right several unfortunate errors which 
occurred in the earlier report on the Scotia Hydroids. There should be substituted 
Hincks for Hinks, Lamarck for Lamark, Johnston for Johnstone, and Halecium for 
Haleciun on p. 523. Aglaophenia dichotoma of the former report I now identify 
with A. heterodonta, Jaderholm (infra, p. 96). 

In the following list, which takes the place of that in the original report, I have 
included all the species collected by the Scottish expedition, those discussed in the 
present paper being indicated by an asterisk : — 



{ 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 67 

GYMNOBLASTEA. 

Family Podocorynid^e. 

Podocoryne carnea, Sars, 1846. 

Family Myriothelid^e. 

* Myriothela austro-georgix, Jaderh., 1904, p. 69. 

Family Eudendrid^e. 

* Eudendrium annidatum (?), Norman, 1864, p. 70. 

Family Atractylice. 

* Perigonimus repens (??), (Wright, 1858), p. 70. 



Halecium arboreum, Allman, 1888. 
(= ,, robustum, Allman, 1888.) 
„ beanii (Johnston, 1838). 



CALYPTOBLASTEA. 

Family Haleciid,e. 



Halecium lialecinum (Linn., 1758). 

,, interpolatum, Ritchie, 1907, (1). 
,, tenellum, Hincks, 1861. 



Family Campanularid^e. 



* Clytia johnstoni (Alder, 1857), p. 71. 
Campanularia angulata, Hincks, 1861. 

,, clytioides (Lamx., 1824), p. 71. 

,, tincta, Hincks, 1861. 

sp., Ritchie, 1907, (1). 

* Obelia geniculates (Linn., 1758), p. 72. 

* ,, hyalina, Clarke, 1879, p. 72. 

* ,, longissima (Pallas, 1766), p. 72. 



* Eucopella crenata(1), Hartlaub, 1901, p. 73. 
Silicularia hemisphxrica (Allman, 1888). 
Hebella striata, Allman, 1888. 

„ „ var. plana, Ritchie, 1907, (1). 

Calycella syringa (Linn., 1758). 

* Campanulina clrilensis, Hartlaub, 1905, p. 74. 

* Thyroscyphus tridentatus (Bale, 1893), p. 74. 



Family Lafoeid^e. 



Lafoea aidarctica, Hartlaub, 1905. 
„ gracillima (Alder, 1857). 
,, ,, var. benthophila, n. var., p. 76. 



Grammaria magellanica, Allman, 1888. 
Brucella armata, Ritchie, 1907, (1). 



Family Sertularid^e. 



Sertularella arborea, Kirchenpauer, 1884. 
„ contorta, Kirchenpauer, 1884. 

„ filiformis, var. reticulata, Ritchie, 

1907, (1). 
„ fusiformis (?), Hincks, 1861, p. 77. 

„ gayi (Lamx., 1821), p. 78. 

,, rectitlieca, Ritchie, 1907, (1). 

tenella (Alder, 1857). 
,, tricuspidata (Alder, 1856). 



* Sertularia cornicina (M'Crady, 1859), p. 78. 

* „ heterodonta, n. sp., p. 79. 

* 

* 



,, mayeri, Nutting, 1904, p. 81. 

,, operculata, Linn., 1758, p. 82. 

,, rathbuni, Nutting, 1904, p. 83. 
* Thuiaria articulata (Pallas, 1766), p. 84. 
(= ,, pectinata, Allman, 1888.) 
Synthecium robustum, Nutting, 1904. 
Staurotheca reticulata, Ritchie, 1907, (1). 



68 MR JAMES KITCHIE : SUPPLEMENTARY REPORT ON 

Family Plumularid^e. 

* Plumularia curvatn, Jiiderholm, 1904, p. 86. * Monostxchas quadridens (M'Crady, 1859), p. 91. 

(= ,, magellanica, Hartlaub, 1905.) * Antenella quadriaurita, n. sp., p. 92. 

,, echinulata, Lamarck, 1836, p. 87. * Aglaophenia allmani, Nutting, 1900, p. 93. 

,, lagenifera, var. septifera, Torrey, 1902, * ,, dubia, Nutting, 1900, p. 95. 

p. 87. * ,, heterodonta, Jaderholm, 1904, p. 96. 

,, pinnata (Linn., 1758). (= ,, dichotoma, of first Scotia Report.) 

„ setacea (Ellis, 1755), p. 89. * „ minima, Nutting, 1900, p. 97. 

,, unilateralism Ritchie, 1907, (1). * ,, latecarinata, Allman, 1877, p. 98. 

Antennularia hartlaubi, Ritchie, 1907, (1). * Halicornaria longicauda, Nutting, 1900, p. 98. 



* 



Antennopsis scotix, Ritchie, 1907 (1), p. 90. 



The localities from which the species recorded in this paper have been obtained are so 
scattered that, for convenience of reference, I have brought them together in list form. 

Station 313, 62° 10' S., 41° 20' W. (S. of South Orkneys). 
Lafo'ea gracillima, var. benthophila, n. var. 

Scotia Bay, South Orkneys. 

Myriothela austro-georgix, Jaderhohn. 

Burdwood Bank, 54° 25' S., 57° 32' W. 

Campanulina chilensis, Hartlaub. 

Port Stanley, Falkland Islands. 

Perigonimus repens (? 1) (Wright). 
Plumularia curvata, Jaderholm. 

Gough Island. 

Obelia longissima (Pallas). 
Thyroscyphus tridentafus (Bale). 
Antenella quadriaurita, n. sp. 
Cape Colony. 

Houtjes Bay (Saldanha Bay). 

Plumularia echinulata, Laink. 
Saldanha Bay, entrance to. 

Eudendrium annulatum (?), Norman. 
Obelia geniculata (Linn.). 
Sertularia operculata, Linn. 
Tlmiaria articulata (Pallas). 
Plumularia lagenifera, var. septifera, Torrey. 
Antennopsis scotim, Ritchie. 
Aglaophenia heterodonta, Jaderholm. 
St Helena. 

Sertularella gayi (Lamx.). 
Station 81, Abrohlos Bank, Brazil, 18° 24' S , 37° 58' W. 
Sertularia cornicina (M'Crady). 
,, heterodonta, n. sp. 
,, rathhuni, Nutting. 
Monostxchas quadridens (M'Crady). 
Aglaophenia allmani, Nutting. 
,, dubia, Nutting. 

,, minima, Nutting. 

Halicornaria longicauda, Nutting. 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 69 

St Vincent, Cape Verde Islands. 

Eucopelln crenata (?), Hartlaub. 

Sertularella fusiformis (1), Hincks. 
Lat. 27° 54' N, long. 33° 17' W. 

Glytia johnstoni (Alder). 

Sertularia mayeri, Nutting. 
Station 537, 29° 54' N., 34° 10' W. 

Campanularia clytioides (Lamx.). 

Obelia hyalina, Clarke. 
Station 538, 32° 11' N, 34° 10' W. 

Plumularia setacea (Ellis). 

Aglaophenia latecarinata, Allman. 

Myriothela austro-georgise, Jaderholm, 1904. 

Several specimens of this bizarre Hydroid have to be recorded. All came from a 
single neighbourhood, Scotia Bay in the South Orkneys, but the depths at which the 
specimens were obtained varied. Some of the examples have already been described by 
Professor J. Arthur Thomson in a short paper in which he regards them, not without 
hesitation, as the separated gonostyles of some unknown giant Siphonopore (Thomson, 
1904). There can be no doubt, however, that these specimens are identical with those 
found by both the Swedish and the French Antarctic expeditions, and recorded by Drs 
Jaderholm and Billard (1906, p. 4) as Myriothela austro-georgice. The length, the 
thickened basal portion on which the blastostyles (each bearing its male or 
female gonophores and a distal tentacle or two) are massed, and, most characteristic 
of all, the capitate tentacles scattered irregularly over the whole hydranth, even 
amongst the blastostyles — these features show that our examples belong to the same 
species as theirs. Nor can there be any doubt that Jaderholm was correct in regarding 
his specimens as belonging to the genus Myriothela, for their resemblance to the 
northern forms is striking, — solitary hydranths, absence of hydrocaulus, capitate 
tentacles scattered over the body, blastostyles grouped at the base of the hydranth, the 
presence of longitudinal folds of endoderm lining the inner cavity. 

Professor Thomson remarks that some of the colonies bore solitary gonophores, while 
one had as many as seven on its blastostyles, and suggests the possibility of the presence 
of two species. Since, however, the specimens examined by Jaderholm had generally 
from one to three, but sometimes as many as six female gonophores, while the male 
gonophores occasionally numbered even ten on a single blastostyle, the variation is so 
great that little stress can be laid on this as a specific character. 

A water-colour sketch made on the capture of one of the specimens indicates that 
their colour was a stronger and brighter orange than is shown by Jaderholm's figure. 
(Jaderholm, 1905, pi. i.) 

Locality. — Scotia Bay, South Orkneys ; dredged in 10 fathoms, April 1903 ; dredged 
in 9 to 10 fathoms, May 1903 ; dredged among mud and pebbles, 18th December 1903. 



70 MR JAMES ETTCHIE : SUPPLEMENTARY REPORT ON 

One specimen was found " on the surface of the water, in a hole which had been cut in 
the ice. The depth of the water at that place was 20 to 30 fathoms ; the temperature 
was 29° F." 

M. austro-georgice has previously been recorded from Cumberland, South Georgia, 
(Jaderholm), and from Flanders Bay and Booth- Wandel Island (Billard). 

Eudendrium annulatum (?), Norman, 1864. 

Two small clumps of bushy colonies appear to belong to this species, but the' 
weathering of our specimens, and the indefiniteness of the characters which differentiate 
the species of Eudendrium, render certainty impossible. The colonies are 5 cm. high, 
and agree with Canon Norman's species in being bushy and beset with very numerous 
branchlets ; in possessing thick, rugged stems, on the surface of which, near the base, 
the fascicular tubes are more or less contorted ; in having branches closely covered with 
strongly marked rings ; and in bearing hydranths with about from 16 to 18 tentacles. 
On the other hand the gonophores, all of which are female, are borne on tentacle- 
bearing hydranths and not on atrophied individuals. So many, however, are the 
gonophores and so closely are they packed around the hydranth, that in not a few cases 
it was difficult to distinguish the presence of tentacles. Since, in some species at least, 
the loss of the tentacles is a degenerative change keeping pace with advancing maturity, 
their presence in this case may be of less significance than at first one tends to regard it. 

These measurements were made : — The diameter of the unfascicled branches and 
branchlets is almost constant, about 0*18 mm. The hydranths are about twice as long 
as broad, the breadth being measured at the level of the bases of the tentacles (0'57 mm. 
long, 0*28 mm. broad). 

Locality. — Entrance to Saldanha Bay, Cape Colony. Depth, 25 fathoms. 21st May 
1904. 

Eudendrium annulatum is a North Atlantic form which has been recorded from 
Shetland (Norman, 1864); Jan Mayen (Marktanner-Turneretscher, 1890) ; Pas-de- 
Calais (Betencourt, 1899) ; Norwegian Coast (Bonnevie, 1899). 

Perigonimus repens (??) (Wright, 1858). 

Scanty material which I have, not without doubt, referred to the above species was 
collected on the shore at Port Stanley. The stems, with a diameter of from , 04 to 
0"05 mm., arise from a stolon creeping upon an encrusting Polyzoon, and reach a height 
of 10 mm. They bear a considerable number of branches which leave the stem at a 
sharp angle and carry secondary, and these sometimes twigs of tertiary degree, in such 
,i way as to give the colonies the appearance of being dichotomously branched. The 
otlslioots can always be distinguished, however, by the presence of a slight constriction 
at their bases. The stems bear distinct rings at their bases and here and there through- 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 71 

out their course, while the intermediate portions are more or less corrugated. A 
delicate chitinous envelope surrounds the coenosarc and is adorned with minute sand 
particles, fragments of sponge spicules, etc. , this coating being continued over the lower 
part of the hydranth. The hydranths, which are in poor condition, appear to be rather 
globular in shape and have from 12 to 15 tentacles. Short-stalked gonophores occur 
scattered over the hydrocaulus, but they are far from mature and offer no characters of 
significance. 

I have no hesitation in identifying my specimens with those collected by Paessler 
at Port Stanley in 1895 and described by Hartlaub (1905, p. 530), although our 
examples bear more numerous branches than his " gar nicht oder nur schwach 
verzweigten Hydrocauli." And I follow Hartlaub, but with considerable hesitation, in 
referring the colonies to the Perigonimus repens of Wright, an almost unbranched form 
with a maximum height of "^ inch" (Hincks, 1868, p. 90), contenting myself merely 
with adding a second mark of interrogation to that which expresses Hartlaub's doubt. 

Locality. — Growing on an encrusting Polyzoon, from seaweed found on the shore at 
Port Stanley, Falkland Islands. January 1903. 



Clytia johnstoni (Alder, 1857). 

Of this common European species only a few stems occur on Saragassum weed. 
They resemble miniature British examples of C. johnstoni, structurally alike in every 
detail, but altogether on a much smaller scale. They are even less in some measure- 
ments than the small variety found by Dr Billard (1907, (1) p. 168) on material from the 
Saragassum Sea. 

Measurements : — 

Stem, length . . . . . .1-6 -2'1 mm. 

„ diameter ..... 0-63-071 „ 

Hydrotheca, length ..... 0-66-07 „ 

,, diameter at margin . . . 0"41-0"48 „ 

Locality. — Off Saragassum weed, from lat. 27° 54' N., long. 33° 17' W. 28th June 
1904. 

Campanularia clytioides (Lamouroux, 1824). 

Several specimens of this minute species have been found creeping on Saragassum 
fronds. The stems are short, measuring from 074 mm. to 0'95 mm. in length and 0'1 
mm. in diameter, and are marked by about eight compact rings at the base of the 
hydrocaulus and about six less compact rings beneath the hydrotheca. The inter- 
mediate portion of the stem is smooth or only slightly corrugated. The hydrothecse are 
short (0*34 mm.), rather broad at the mouth (0*38 mm.), and taper rapidly to the base. 
Their walls are thick, but vary considerably in different individuals and even in the 
various parts of the same individual. At the margin of the shelf which divides the 



72 MR JAMES RITCHIE : SUPPLEMENTARY REPORT ON 

cavity of the hydrotheca proper from the small globular cavity at its base is a ring of 
bright dots, indicating the points at which the base of the polyp was attached to the 
hydrothecal wall. 

The gonosome is not present. 

Locality.— On gulf weed from Station 537. Lat. 29° 54' N., long. 34° 10' W. 
29th June 1904. 

Obelia geniculate! (Linnaeus, 1758). 

In addition to the Gough Island locality given in the former report, this widely 
distributed species occurs from another station. Gonophores are present on the colonies 
from habitat (/>). 

Locality. — (a) Entrance to Saldanha Bay, Cape Colony. Depth. 25 fathoms. 21st 
May 1904. (b) Growing plentifully, along with Plumularia lagenifera, var. septifera, 
on the segments and telson of a lobster, Palinostus lalandii, from the same locality. 

Obelia hyalina, Clarke, 1879. 

Many small, 6-mm. high colonies of this species occur creeping over the fronds 
of Saragassum weed in association with Campanularia clytioides. The characters 
are as described and figured by Clarke, and the dimensions of our specimens 
appear to agree more closely with those of the type than do the corresponding 
measurements of the large-sized Talisman examples described by Dr A. Billard 
(1907, p. 170). 

Measurements : — 

Hydrotheca, depth . . . . 0'2 mm. 

,, maximum diameter . . . 0'18 ,, 

Peduncles, length ..... "36-0 '45 mm. 

Stem internodes, length .... 0'54 mm. 

No gonangia were present. 

Locality. — Creeping on gulf weed from Station 537. Lat. 29° 54' N., long. 34° 10' W. 
29th June 1904. 

Obelia longissima (Pallas, 1766). 

In addition to the colonies already recorded from the South Orkney Islands, a small 
fragment whose characters agree with those of the above species has been found at Gough 
Island. 

Measurements : — 

Hydrotheca, depth ..... - 38-0'45 mm. 

„ diameter at margin . . . 0'38-0'45 ,, 

Locality. — Off Gough Island. Depth, 25 fathoms. Bottom, rock. 23rd April 
1904. 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 73 



Eucopella crenata (?), Hartlaub, 1901. 

A very few examples, whose occurrence has already been mentioned (Ritchie, 1907, (2) 
p. 488, footnote), were creeping on seaweed obtained on the shore of St Vincent, Cape 
Verde Islands. The absolute identification of these examples with Professor Hartlaub's 
E. crenata is, in the absence of the gonosome, impossible, and, considering differences 
which occur in the shape of the hydrotheca and in the ringing of the peduncle, 
appears to me to be a matter of doubt. Our specimens, however, are identical 
with those described by Dr A. Billard, 1907, and I follow him in assigning them 
(with a query) to the above species. The hydrotheca are more conical than 
those of the type, and their proportions differ slightly from those of Billard's 
specimens, where the depth exceeds the breadth, for here these dimensions are 
identical, or the latter may even exceed the former. The margin of the hydrotheca 
is cut into about thirteen rounded teeth, and above the projecting septum at the 
base there occurs a circle of from 28 to 31 bright dots (indicated in Hartlaub's fig. 27). 
The peduncles bear from 9 to 12 sharply defined rings at the base, and from 1 to 4 
annulations at the summit, while the median portion is generally more or less 
corrugated. 

Measurements : — 

Length of peduncle ..... 1*01-1 "98 mm. 

Breadth „ ..... 0*07-0-09 „ 

Depth of hydrotheca ..... 0*32-0*45 „ 

Breadth „ at margin . . . 0*34-0*40 ,, 

Locality. — Growing on seaweed from the shore to the N.E. of Porto Grande, 
St Vincent, Cape Verde Islands. 1st December 1902. 

Apparent evidences'of regeneration are given by some of the specimens. To mention 
one example. The hydrotheca is borne on a stem 1'68 mm. in length, but this stem is 
composed of two distinct portions, a basal section 1*04 mm. long with walls 9m in 
thickness, of a dark horn colour and deeply stained by eosin. This portion is 
sharply truncated at the distal end. It is succeeded by a shorter section 0'64 mm. 
long, with walls only 6m in thickness, clear and transparent and faintly tinged 
with eosin. It is apparent that the peduncle has been truncated at a particular point, 
and that from this point the regenerating hydranth has built a new perisarcal 
protecting sheath which, as one would expect, differs in density from the older 
portion. It is interesting to note that the regenerated portion has, instead of 
continuing the smooth character of the old peduncle, assumed the character of a 
complete peduncle, for it bears four clean-cut rings above its point of origin, these 
being followed by a corrugated portion, and this by three rings below the hydrotheca. 
In all the cases examined the regeneration process has reproduced a complete but 
miniature peduncle. 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 4). 10 



74 MR JAMES RITCHIE : SUPPLEMENTARY REPORT ON 

Campanulina ckilensis, Hartlaub, L905. 

Three minute colonies epizoic on Halecium beanii belong to this species. The 

ringing and branching of the stems and the shape of the hydrothecse agree with the 

descriptions and figures of Hartlaub and of Jaderholm (1905), but although the former 

says of the hydrothecse that they are " viel weniger conisch als die von C. repens, 

Wright" [ ! Allman, 1864], I cannot distinguish between the hydrothecse of the South 

American and British species. C. cltilensis is, however, more frequently branched than 

C. repens, but I doubt whether this and the other rather indefinite characters cited by 

Hartlaub are constant and of specific value. Scarcity of material of both the South 

American and British forms, however, prevents a comparison sufficiently extensive to 

warrant me in setting them down as belonging to the same species. 

Measurements : — 

O. ckilensis. 0. repens* 

Stem, diameter . . . 0'06 mm. 0'08 mm. 

Hydrotheca, length . . . 0-20-021 mm. 0-18-0-28 mm. 

,, diameter at mouth . - 09 mm. - 10 mm. 

Locality. — Growing upon Halecium beanii from Station 346, Burdwood Bank. 
Lat. 54° 25' S., long. 57° 32' VV. Depth, 56 fathoms. 1st December 1903. 

Jaderholm has already recorded C. ckilensis from Burdwood Bank, where it was 
growing on the carapace of a crab at 137-150 metres. Billard (1906, (2) p. 12), found 
the species in collections from Flanders Bay and Booth- Wandel Island. Hartlaub's 
specimens were obtained at Calbuco on Tubularia and Eudendrium. 

Thyroscyphus tridentatus (Bale, 1893). 

A few colonies have been found at a single locality. Simple, unbranched stems, 
1 mm. in height, arise from a creeping hydrorhiza at intervals of some 2 or 3 mm. To 
the unaided eye the colonies have an erect, rigid appearance, and are seen to bear pro- 
minent hydrothese placed alternately on the stem. Under the microscope the stem 
resolves itself into a series of distinct internodes which vary considerably in size, a much 
shorter being occasionally wedged in between two longer individuals. That the nodes 
in our specimen are apparently more distinct than those in the examples recorded by 
Professor Hartlaub (1901, p. 369) from French Pass, to the north of the South Island of 
New Zealand, is of little importance, as the boundary mark between internodes is liable 
to considerable variation. The locality of the node is, moreover, rendered more evident 
in the Scotia specimens, because the proximal end of each internode is generally 
marked by a rude annulation. The hydrothecse are arranged alternately, and the 
whole series lies in one plane. 

A hydrotheca rests upon a short process at the distal end of each internode. The 
hydrotheca is separated by a distinct boundary line from the internodal process, and 
occasionally one, two, or even three short joints intervene between them. The distal 

* Specimens from Plymouth in my collection. 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 75 



margin of such an interpolated joint is always abrupt, and the perisarc of the existing 
calycle is not directly continuous with this margin, but is found to merge with that of 
the interpolated joint some distance within the joint (see fig. 16). From this it would 
appear that, the joints are not all contemporaneous, but represent the remains of former 
cups which, broken off by accident, have been replaced once, twice, or oftener by the 
regenerative power of the ccenosarc. In this case, therefore, they can be of no diagnostic 
value. Evident knobs of chitin are present on the inner surface of the adcauline wall 
at the base, and of the abcauline wall at the margin. The former do not seem to be 
reproduced in regenerated hydrothecae. 

The hydrothecse are much more deep than broad, and are characterised by an almost 
straight abcauline and a strongly convex adcauline contour. They are in most cases 
considerably longer than the stem internodes. The margin is divided into three pro- 





a. b. 

Fig. 1. — Thyroscyphus Iridentaius. x 70. (a) Primary liydrotheca showing l\ and k 2 , knobs of chitin ; pp, points on 
hydrotheca to which the hydranth is moored by ccenosarcal strands. (6) Internode process and regenerated hydrotheca ; 
Jc, knob of chitin marking base of primary liydrotheca ; w , wall of original hydrotheca ; w lt wall of first regenerated 
hydrotheca ; w„, wall of second regenerated hydrotheca. 

nounced teeth, separated by three deep and graceful bays, and is furnished with a three- 
flapped operculum. No trace of a gonosome could be found. 

The following measurements give some indication of the variable proportions of our 
specimens : — 

Stem, length . 
Internode, length 
,, breadth 



Hydrotheca, length 

,, breadth (maximum) 



Up to 13 mm. 
0'49*-0-91 mm. 
0-13-0-22 mm. 
060-0-63 „ 
0-22-0-28 „ 



Locality. — Gough Island. Depth, 25 fathoms. Bottom, rock. 23rd April 1904. 

Distribution. — -The species was originally described by Bale from material obtained 
at Port Phillip, in the south of Australia, as Campanularia tridentata. Subsequently 
it has been recorded by Professor Hartlaub (1901) from French Pass, north of South 
Island, New Zealand. These, so far as I am aware, constitute the only records of the 
species, so that the Scotia specimens from Gough Island, midway between Cape of Good 
Hope and Cape Horn, extend its known range from the South Pacific to the South Atlantic. 

* A regenerated internode which succeeded an old truncation of the stem. 



76 



MR JAMES RITCHIE : SUPPLEMENTARY REPORT ON 



Lafo'ea gracillima (Alder, 1857), var. benthophila* n. var. 

This species lias to be recorded from an additional locality, to the south of the South 
Orkney Islands. The two specimens from this locality are, however, of a type quite 
distinct from the Burdwood Bank examples. They are both small and incomplete. 
The larger, 18 mm. high, bears two short branches ; the smaller, 9 mm. high, is branch- 
less. Only a trace of fasciculation is exhibited by the latter, but the stem of the former 
is a typical rhizocaulom. The hydrothecse, while they closely resemble those of typical 
examples of L. gracillima, differ in being placed at more regular intervals on the stem, 
in preserving a near approach to alternation, in lying towards the stem at a much 
smaller angle, and in lacking a twist on the hydranthophore. 




Flo. 2. — Lafoea gracillima, var. benthophila. Portion of branch with hydrothecse. x 30. 

I do not lay much stress on this last point, however, since untwisted hydranthophores 
are not unfrequently present in typical colonies of the species. The most striking 
difference lies in the robustness of the hydrothecse, the dimensions being much greater 
than in any other specimen I have examined. Near the base of the hydrotheca proper 
is a rude row of minute, refringent prominences on the internal wall, to which the 
base of the polyp was presumably attached. Occasionally, but only where the margin is 
reduplicated, two rows of dots occur. 

The following table indicates in millimetres the differences, in respect of size, 
between the present and typical specimens : — 





Coat's Land 
Specimen. 


Burdwood Bank 
Specimen. 


North Sea 
Specimen. 


Eydrotheca, including hydranthophore 

,, diameter at mouth 
Diameter of a simple tube .... 


0-87-101 

021-0-25 

016 


0-63-0-64 

0-11-0-13 

0-11 


057-0-76 
011 

o-io 



* PevOos, the deeps ; and </>i'aos, loving. 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 77 

No gonangia were present. 

Locality.— Station 313. Lat. 62° 10' S., long. 41° 20' W. Depth, 1775 fathoms. 
Bottom deposit, blue mud and boulders. 18th March 1903. 

The distribution of the species is almost world-wide, but it has not hitherto been 
recorded from Antarctic Seas. 

Sertularella fusiformis (?) Hincks, 1861. 

A few minute simple stems, 7 mm. high, rising from a stolon creeping upon a sea- 
weed. The slightly geniculate stems are divided into internodes which vary consider- 
ably in length, those nearer the base being longer than the more distal, the former 1*08 
mm. as against an average of 0'61 mm. for the latter. The internodes are narrow, only 




Fig. 3. — Sertularella fusiformis (?) Portion of stem, x 40. 

0"06 mm. in diameter at the base, but they widen upwards until a ledge is formed upon 
which the base of the hydrotheca rests. They are separated by slanting nodes, immedi- 
ately above which occur one or two rings, while the remainder of the internode is more 
or less definitely wrinkled. 

The hydrothecae lie in the same plane, are alternate, and are placed one on the distal 
end of each internode. Rather more than half of each is free, the free portion leaning 
well away from the stem. In shape they are flask-like, bulging proximally, while towards 
the distal end there is formed by a sudden constriction a distinct " neck," which is sur- 
mounted by the four-toothed margin. The walls of the hydrothecse are smooth 
externally, but just within the margin the inner surface bears four distinct blunt teeth 
which lie midway between the marginal teeth. 



78 MR JAMES RITCHIE : SUPPLEMENTARY REPORT ON 

Dimensions of hydrotheca :- — 

Length ...... 0-45-0-52 mm. 

Diameter where it becomes free from internode . - 25-0"27 ,, 

Diameter of " neck " ..... 014-0-16 „ 

No gonangia were present. 

It is with some doubt that these specimens, their gonangia lacking, have been referred 
to the S. fusiformis of Hincks. The hydrothecse in our specimens appear to be more 
robust, to have a more decided " neck," and a more pronounced inclination away from 
the stem, while the presence of internal teeth is not mentioned in Hincks's description. 
The present specimens closely approach the very doubtful S. fusiformis (??) described by 
Professor Cl. Hartlaub (1900, pi. 5, fig. 9) from Eovigno on the Adriatic Sea, but the 
ringing which is absent in his is markedly present in our examples. 

Locality. — Growing on seaweed found on the shore to the N.E. of Porto Grande, 
St Vincent, Cape Verde Islands. 1st December 1902. 

Sertularella gayi (Lamouroux, 1821). 

In addition to the colony found near Gough Island, another of almost equal dimen- 
sions, 12 cm. high, has to be recorded from St Helena. Its minute structure is very 
similar to that of the Gough Island example, the rugosities on the upper surface of the 
hydrothecse being in some cases almost obsolete. 

This species, although widely distributed in the Northern Atlantic and in the 
Mediterranean Seas, has hitherto been recorded south of the equator only from the Cape 
of Good Hope (Algoa Bay), (Hartlaub, 1905, p. 613). In conjunction with this occur- 
rence the two Scotia records, from Gough Island and St Helena, may be taken to indicate 
a wide distribution for the species in the South as well as in the North Atlantic. 

Locality. — Intertwined with an Alcyonarian, Amphilaphis regularis, from St Helena. 
30th May 1904. 

Sertularia comicina (M'Crady, 1859). 

Scanty material, a mere half-dozen colonies, represent this species. The stems arise 
from a stolon creeping upon the surface of the fragment of Codium also invested by 
& rathbuni and S. Jieterodonta, from the former of which, indeed, they are almost 
indistinguishable to the unaided eye. The largest is only 6 mm. in height. The 
specific characters agree with those given by Nutting (1904, p. 58), but the following 
variations were noted. The stem internodes are proportionally longer than in 
Nutting's specimens, for while in his examples the " height of the hydrothecse is 
usually about equal to that portion of an internode which lies between the hydrothecal 
base and the node below," in ours the latter distance considerably exceeds the former. 
Our hydrothecse, again, have a longer distal portion free from the stem, the result being 
that the proportion of the anterior adnate part to the whole length of the hydrotheca is 



THE HYDKOIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 79 

reduced from " about two-thirds " to about one-half. As seen from the anterior aspect 
the hydrotheca appears to be covered in by only two opercular flaps, the free margins 
of which run from the tip of one lateral tooth to that of the other ; but when a 
lateral view of the colony is obtained, so that one can look directly on the operculum, 
another line is seen running from the mid point of the horizontal line connecting the 
lateral teeth to the median superior point of the aperture. This line presumably marks 
the margin of two contiguous flaps of the operculum, and it would therefore seem that 
three, and not two, are present. The determination, however, is one of considerable 
difficulty owing to the delicacy of the operculum. The downward projecting processes 
from the base of the hydrotheca are long and evident. 

Measurements : — 



Internodes, length 

,, breadth 

Hydrotheca, length of contiguous portion 
„ ,, free portion * 

,, diameter near base 

„ at aperture 



0-63-0-78 mm. 
0-05-0 06 „ 
0-22-0-25 „ 
0-21-0-22 „ 
0-08-0-09 „ 
0-08-0-09 „ 



Locality. — Growing upon seaweed (Codium, sp.) from Station 81, Abrohlos Bank, 
Brazil. Lat. 18° 24' S., long. 37° 58' W. 36 fathoms. Bottom deposit, coral. 20th 
December 1902. 

Sertularia heterodontaj n. sp. 

Growing upon the surface of a seaweed (Codium, sp.) in company with S. rathbuni 
and & cornicina are numerous colonies of this more minute form. The simple 
unbranched stems arise from a creeping stolon and cover the weed as with a coat of 
delicate hairs. At first glance the colonies are hard to distinguish from those of 
S. rathbuni, but on close examination their minuteness and delicacy render them 
distinct even to the unaided eye. The largest are only 6 mm. in height, 3 - 5 to 4 mm. 
being much more common. The hydrorhizal tube has a diameter a little greater than 
that of the stem, and is strengthened in places by chitinous processes projecting down- 
wards from its roof or upwards from its floor. 

The stems are divided into fairly regular internodes separated by distinct nodes. 
At the base two oblique nodes occur in succession, cutting off between them a short 
lozenge-shaped athecate internode. The remainder of the internodes are long and 
slender, narrower at the proximal end and very gradually increasing in diameter 
upwards, until by a sudden dilation at the top they form a broad bracket upon which 
the hydrothecse rest. The thecate internodes are separated by straight nodes, but 
occasionally an oblique node occurs in addition, a small athecate internode beino- in 
such a case intercalated between the longer thecate individuals. 

* The " free portion " is measured from the line of the stem to the tip of the lateral teeth. 

t eVtpos, other ; and <Wos, a tooth, indicating the presence of other than the usual marginal teeth. 



80 



MR JAMES RITCHIE : SUPPLEMENTARY REPORT ON 



An opposite pair of hydro thecse rests on each internode, six to seven or even ten 
pairs being found on one colony. The hydrothecee are set slightly on front of the stem, 
are always contingent in front for rather more than one-third of their total length, but 
remain separate behind. The free portion diverges abruptly at a wide angle. Viewed 
from the anterior aspect the sides of the adherent portion of a hydrotheca are parallel 
to the long axis of the stem, while those of the free portion converge towards the 
aperture. The contour lines, especially in the younger hydrothecae, are remarkably 
straight and graceful. Beneath the innermost angle of the base there is a minute 
chitinous thickening, and another projects from the base into the interior of the 
hydrotheca. The latter appears to be roughly triangular in shape, the apex pointing 






Fig. 4. — Sertularia heterodonta. (a) Anterior aspect of stem, x 60. (b) Lateral aspect of stem showing three-flapped 
operculum, x 60. (c) Hydrothecse with reduplicated margins, one regenerated margin bearing internal teeth, x 100. 
(d) Gonotheca. x 50. 

into the cup, the side towards the centre of the stem being thickened and concave, 
that remote from the centre sloping gradually till it merges with the hydrothecal floor. 
This process represents a small ridge bounding the posterior edge of the opening 
between hydrotheca and stem. The margin of the hydrotheca is divided into three 
distinct and sharp teeth, the lateral pair the more prominent, the median and superior 
individual tilted slightly upwards. There are three opercular flaps, difficult to 
distinguish except when one is looking directly into the aperture of the hydrotheca. 
Just within the margin and projecting from the inner wall are three prominent internal 
teeth, triangular in shape, with blunt apices tilted towards the aperture. These 
alternate with the marginal teeth. In hydrothecse towards the base of the stem the 
superior pair appears occasionally to be undeveloped. Frequently the free portion of 
the tube is much elongated by the reduplication of the margin, and in some cases a 
new series of internal teeth is produced on the added portion (fig. 4, c). 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 81 

The gonangium, of which I have been able to find only one example, is borne on 
the front of the stem immediately beneath a hydrotheca. It is oval in shape, tapering 
below to form a distinct stalk, and truncated above by a wide aperture. This is 
bounded by an insignificant thickened neck punctuated by small refringent chitinous 
swellings on the inner surface. The wa]ls of the gonangium are smooth. 

Measurements : — 



Intern odes, length 




0-46-0-62 mm 


,, diameter near base 




003-0-04 „ 


Hydrotheca, length of contiguous 


portion 


0-10-0-14 „ 


„ ,, free portion 


* . 


0-14-0-15 „ 


,, diameter at base . 




0-06-007 „ 


„ ,, „ aperture 




004 mm. 


Gonangium, length 


. 


0-74 „ 


,, maximum diameter 




0-41 „ 



One colony shows a pair of hydrothecse at an early stage of development, ere yet 
the aperture, or the base, or the internal teeth had been formed. The stolon has so 
regulated itself to the papillated surface of the Codium upon which it is growing that 
the outline of its under surface is made up of a series of crescentic bays. 

The three-toothed margin, the three-flapped operculum, the prominent internal 
teeth, the presence of a projection from the base into the interior of the hydrotheca, 
together with the straight contours of the hydrothecse and the length of the internodes, 
are points which distinguish this species. It has affinities with S. linealis, Warren, 
1908. 

Locality. — Growing upon seaweed from Station 81, Abrohlos Bank, Brazil. Lat. 
18° 24' S., long. 37° 58' W. Depth, 36 fathoms. Bottom deposit, coral. 20th 
December 1902. 

Sertularia mayeri, Nutting, 1904. 

Small colonies, 5 mm. high — less than half the height of the type specimens— occur 
creeping on Saragassum weed. Their characters agree with Nutting's description and 
figures, and are identical with those of specimens collected by Mr Crossland at the 
Cape Verde Islands (Ritchie, 1907, (2) p. 505), although, owing to their smaller size and 
the consequent restriction of the number of hydrotheca-pairs (in the largest specimen 
mounted for microscopical examination there are only six pairs), the variation between 
the proximal and distal pairs is not so marked as in the larger Cape Verde specimens. 
Even here, however, the distinction between the distal long, narrow, closely forked, 
gradually separating pairs, and the proximal short, dumpy, wide-spreading pairs is 
sufficiently distinct and characteristic. Tt is indicated by the measurements below. 
In the majority of the calycles two lateral teet'h and a smaller median and superior 
tooth, which succeeds an indentation in the wall of the hydrotheca and curves slightly 
upwards, are very evident. 

* From the line of the stem to the tip of the lateral teeth. 
TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 4). 11 



82 



MR JAMES RITCHIE : SUPPLEMENTARY REPORT ON 



Measurements (a and b are distinct colonies) :- 





Basal. 


Distal. 


Length of internode .... 
,, liydrotheca .... 
Breadth of hydrotheca-pair from tip to tip 

„ ,, „ at base 


a. b. 
0-42 ; 0-49 mm. 
0-29; 0-30 „ 
0-48; 0-46 „ 
0-22; 0-25 „ 


a. b. 
0-56 ; 0-66 mm. 
0-38; 0-45 „ 
0-56; 0-52 „ 
21; 020 „ 



Locality. — Creeping on gulf weed, lat. 27° 54' N.. 
1904. 



long. 33° 17' W. 28th June 



Sertularia operculata, Linnaeus, 1758. 

Several small fragments of this species occur intertwined with other Hydroids from 
the same locality. The specimens are typical in branching and in minute structure, 
but a considerable amount of variation occurs not only in the length of the two 
hydrothecal teeth, but also in their prominence relative to each other (cf. Hartlaub, 
1905, pp. 665, 666). Of the specimens figured by Professor Hartlaub our examples 
most nearly approach those from West Patagonia collected by F. P. Moreno ; but from 
those they differ in the proximity of the hydrothecse, for the teeth of one may reach 
the level of the base of its successor. The present examples are also characterised by 
the exceedingly minute portion of the distal extremity of the liydrotheca, which is free, 
the proximal side of the aperture lying almost against the internode. 

The points above refer specially to the younger branches. On the older portions 
the hydrothecse are only sub-opposite, their length is less relatively to that of the 
internode, while a slightly longer distal portion is free. 

A few typical gonangia occur on the branches. They exhibit a tendency to 
asymmetry, the aperture lying towards the outer side of the axis of symmetry. 

The following measurements indicate the relations of the various parts : — 



Length of branch internode 
,, stem internode 
„ liydrotheca . 

,, teeth 

Distance from tip to tip of a 



0-33-0-50 mm. 
0-54-0-61 „ 
0-20-0-25 „ 
0-06-012 „ 
0-43-0-58 „ 



lydrotheca-pair 

Locality. — Dredged at the entrance to Saldanha Bay, Cape Colony, in 25 fathoms. 
2 1st May 1904. 

The distribution of this species is world-wide. In addition to its European localities 
it has been recorded from the coasts of North and South America, of Southern Asia, of 
Australia and New Zealand, and of Africa, although the records from the last continent 
are few. The African localities other than the Scotia record are as follows : — 

South Africa (Busk, 1850); Cape of Good Hope (Eugenie Expedition, Jaderholm, 
1903); Port Natal (Professor J. A. Wahlberg, Jaderholm, 1903); Mauritanian Coast 
(Billard, 1906 (1) ); North- West of Cape Blanc, Soudan (Billard, 1906 (1) ). 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 



83 



Sertularia rathbuni, Nutting, 1904. 

The specimens occur plentifully on a seaweed (Codium, sp.), standing erect and 
rigid on its surface, so that it appears as if covered with a sparse coating of delicate 
hairs. They are simple, altogether without branches, herein differing, but immaterially, 
from the specimens described by Allman (1877), Versluys (1899), and Nutting, some 
of which bore pinnules, and they attain a height of only 10 mm. The structures of the 
stem are as previously described. The hydrothecae are in opposite pairs, the individuals 
of which, in the distal part of the colony, are contingent for almost half their height, 
but the line of contact gradually decreases until in the proximal pair the individuals 






Fig. 5. — Sertularia rathbuni. (a) Anterior aspect of stem, x 50. (b) Three-quarters view of hydrotheca showing 
three- flapped operculum, x 60. (c) Gonangium. x 70. 

may be quite apart. Behind the stem the hydrothecae are always separate. Beneath 
each hydrotheca, at the corner where base and inner wall meet, are two chitinous 
processes which project downwards and lie alongside the wall of the internode. These 
processes are more distinct in the older hydrothecae. The margin of a hydrotheca is 
cut into three teeth, the lateral pair longer than the single superior tooth. There are 
three opercular flaps. 

Gonosome. — The gonangia, which have not previously been described, occur on 
many of the colonies. They are borne on the stem, from one to three in number, and 
arise immediately beneath hydrothecae towards the base of the colony. In the speci- 
mens which I have examined only one gonangium is apportioned to each hydrotheca- 
pair. The gonangia are broadly ovate, marked with about six rather indefinite annula- 
tions. Proximally they taper into a short stalk, while distally they contract into a 
short wide neck, ornamented at its base by a circlet of bright spots, thickenings of the 



84 



MR JAMES RITCHIE: SUPPLEMENTARY REPORT ON 



perisarc. The contents of the gonangia were frequently missing, and in no case were 
they in a state fit for minute examination. 

Measurements : — 



Intern odes, length 

,, breadth 

Hydrotheca, portion free 

,, ,, fixed 

„ diameter at mouth 

Gonangium, length 

,, maximum breadth 



0-81-0-94 mm. 
0-10-0-12 „ 
0-18-0-20 „ 
0-27-0-28 „ 

0-74 mm. 
0-60-0-63 „ 
0-45-0-53 „ 



In general build, in possessing chitinous projections from the base of the hydrotheca, 
and in the structure of its gonangium, this species bears close resemblance to S. cornicina 
(M'Crady) as described by Nutting. The latter species, however, is to be distin- 
guished by the tubular shape of its hydrothecse, by the number of the marginal teeth 
and of the opercular flaps, by the narrowly oval outline of its gonangium, and by the 
fact that the gonangia are borne on hydrorhizal tubes at the base of the colony, and not 
on the stem. 

Locality. — Growing on a Plumularian, Halicornaria longicauda, and on seaweed 
(Codiwn, sp.), both from Station 81, Abrohlos Bank, Brazil. Lat. 18° 14' S., long. 37° 
58' W. Depth, 40-50 fathoms. Bottom deposit, coral. 20th December 1902. 

Previously recorded only from the Gulf of Mexico : Allman, 1877 ; Versluys, 1899 
(Dry Tortugas) ; Nutting, 1904, lat. 29° 28' N, long. 87° 56' W. 



Thuiaria articulata (Pallas, 1766), { — T. pectinata, Allman, 1888). 

In the earlier report, a colony of this species was recorded under Allman's name of 
T. pectinata. The occurrence of an additional colony with gonangia reopened the ques- 
tion of nomenclature, and an examination of Allman's type was made (through the kind- 
ness of Mr R Kirkpatrick, of the British Museum). The examination assures me of 
the identity of T. pectinata, Allman, with T. articulata, Pallas. 

The larger of our two colonies was some 6 cm. high, the height of Allman's speci- 
mens being also " between two and three inches." The colonies are simply pinnate, with 
a monosiphonic stem about 1 mm. in diameter at the base. The stem is partitioned into 
regular and well-marked internodes, each bearing three pairs of opposite hydrothecse, 
and from between the proximal and median pairs arises a pair of opposite pinnae. 
Proximal to the first pinna-bearing internode a few destitute of offshoots occur, and on 
these the number of hydrothecse is not constant. Apart from these only one internode 
has been observed in which three pairs of hydrothecse have not occurred ; and it is 
clearly abnormal, for it lacks pinnae, has but one pair of hydrothecse, and is so short 
that the hydrothecse project beyond it, free for nearly half their height. The pinnae 
stand out from the stem at a wide angle, and they too are divided into distinct but 
less regular internodes, each bearing from two to five pairs of sub-opposite hydrothecse. 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 



85 



(Marktanner-Turneretscher (1890) gives the variation as from three to ten pairs 
per internode.) 

The hydrothecse are almost wholly immersed, and the " free membranaceous extension 
of the wall," too prominently figured in Allman's account, has been destroyed (as indeed it 
was in the T. pectinata specimen examined), leaving a jagged edge level with the general 
outline of the pinna. There is evident on the abcauline wall of the hydrothecse, just 
within the opening, a small knob of chitin, and immediately above or upon this rests the 






PlG. 6.— Thuiaria articulata. (a) Single internode of stem with proximal internodes of pinnae, x 20. (b) Gonangium. 
x 12. (c) Internode from T. pectinata with hydrothecse slightly apart, x 20. {d) Internode from T. pectinata with 
congested hydrothecse. x 20. 

base of the one-flapped operculum. On the stem internodes and on the younger pinnae 
the hydrothecae are slightly apart from each other, but in the older pinnse they are 
compressed and lie closely packed, the distal end of one forced against the base of its 
successor (cf. fig. 6, c and d). 

The gonangia are clustered on one face of the stem and on the corresponding 
faces of the pinnse. They arise immediately beneath a hydrotheca, are elono-ate 
oval in shape, with a wide, circular, distal opening bordered by a distinct neck, 
and with a tapering proximal end. Their distal half bears more or less indefinite 
annular rugosities. 



86 



MR JAMES RITCHIE : SUPPLEMENTARY REPORT ON 



Measurements : — 





Scotia 
Specimens. 


Allman's 
T. pectinata. 


Stem internodes, length 
Hydrothecae on stem, length 

„ ,, breadth 

,, on pinna, length 

,, ,, breadth 
Gonangia, length .... 

,, greatest diameter. 


2-25 mm. 

0-50 „ 

0-17 „ 

045 „ 

0-15 „ 

3 

1-5 „ 


2-75 mm. 
0-52 „ 
0-20 „ 
0-48 „ 
0-18 „ 

> not present. 



Allman's specimens are somewhat more robust than those collected by the Scotia. 

Locality. — Dredged at the entrance to Saldanha Bay, Cape Colony, at a depth of 
25 fathoms. 21st May 1904. 

The species has been recorded from Algoa Bay and Cape of G-ood Hope (Kirchen- 
pauer, 1884) ; South Africa (Busk, 1850) ; Simon's Bay, Cape of Good Hope (Allman, 
1888). 

Plumularia curvata, Jaderholm, 1904 ( = P. magellanica, Hartlaub, 1905). 

Dr Jaderholm has kindly drawn my attention to the fact that the species recorded in 
my earlier report on the Scotia collections as P. magellanica, Hartlaub, had been pre- 
viously described by him as P. curvata, which name I therefore substitute for Hart- 
laub' s synonym. A variation which does not seem to have been present in the examples 
examined by Jaderholm or Hartlaub was described and figured in the earlier report 
(pi. iii., figs. 1, la) ; here I wish simply to add that that variation is more general in the 
hydroclades than I had at first supposed. My previous description reads: — "In the 
proximal hydroclades .... two processes arise below and at the opposite sides of the 
first hydrotheca, each of which bears a thecate internode, so that after the first hydro- 
theca the hydroclade possesses two diverging branches each similar to the simple 
distal hydroclades" (1907, (1) p. 541). But this duplication of the hydroclade occurs 
not only at the first hydrotheca but sometimes at successive hydrothecse as well. From 
beneath the first hydrotheca two diverging internodes spring, each capped by its 
hydrotheca ; from the bases of each of those second pairs arise, and from these again, 
and so on in a system comparable to the false dichotomy of the Mistletoe, until 
dichotomous pairs of even the fourth degree may be reached. Some of the hydroclades 
thus assume a complicated and much-branched appearance, quite distinct from the simple 
type figured by Jaderholm (1905, pi. 14, fig. 10) and Hartlaub (1905, p. 684, fig. N 5 ). 

Measurements : — 



Stem internode, length 


0-42-0 52 mm 


,, ,, breadth 


0-11-0-13 „ 


Hydroclade internode, length . 


0-24-0-23 „ 


Hydrotheca, depth .... 


o-io 


„ diameter at mouth 


0-15-0-17 „ 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 87 

Locality. — " Growing on a sponge, Port Stanley, Falkland Islands. 3rd February 
1904." 

Previous records are from Port Louis and Port Albemarle, Falkland Islands (Jader- 
holm, 1905); Southern Tierra del Fuego, and Island Picton in the neighbouring 
archipelago (Hartlaub, 1905). 

Plumularia echinulata, Lamarck, 1836. 

In addition to a previously mentioned occurrence at Cape Town, a second locality, 
also in Cape Colony, has to be recorded, namely, Saldanha Bay. The specimens from 
this place, while rather smaller in size than the Cape Town examples, are similar in 
minute structure, except that in the former the hydrotheca-bearing internodes are rather 
shorter, and the hydrothecae therefore more congested, than in the latter. 

The gonangia are elongate oval, considerably longer in proportion to their diameter 
than those figured by Hincks (1868, pi. lxv., fig. 26), and possessing shorter and 
more regularly arranged spines. They stand out from the median aspect of the stem 
in a densely packed row. 

Locality. — Shore, Houtjes Bay, Saldanha Bay, Cape Colony. 19th May 1904. 

Plumularia lagenifera, var. septifera, Torrey, 1902. 

Whereas typical specimens of P. lagenifera are about three inches long, are some- 
times branched, and grow in flexuous clumps, the specimens which I have referred to 
Torrey's variety are short (only 7 mm. high), never branched, and are markedly rigid 
in habit. The detailed structure is that of a compressed P. lagenifera, where the inter- 
nodes have become shorter and comparatively stouter, while the internal septa have 
become more distinct. In conjunction with the general shortening it has come 
about that in the intermediate internodes there is generally but one septum, 
although our specimens differ from those described by Torrey — where " no intermediate 
internode has more than one septal ridge " — in that, in several, there are traces of a 
second ridge on the distal side of the nematophore, while in at least one case the second 
ridge is quite pronounced. Torrey is equally emphatic that " there is never more than 
one internode between thecate internodes," but I have observed a case in which two 
successive athecate internodes occurred, the distal being much the shorter and lacking 
a nematophore. There was no evidence that this duplication was due to abnormal 
growth, such as regeneration. These variations, however, only show more clearly the 
relationship between this form and P. lagenifera type, and confirm Torrey's placing of 
it as a variety of that species. 

In one point the Scotia specimens differ both from the type and from the variety, 
for they show no trace of a nematophore on any internode " on side opposite branch 
[i.e. hydroclade] and immediately distal to the proximal septum." 



88 



MR JAMES RITCHIE: SUPPLEMENTARY REPORT ON 



Between the hydroclade and the internode process on which it is set occur from one 
to three athecate internodes. The hydrorhizal tubes are close-set, are compressed from 
above downwards, and are supported by thickenings of the perisarc which project into 
the interior of the tube. Somewhat similar thickenings I have already seen in the 
hydrorhiza of a species of Podocoryne (Ritchie, 1907, t2) p. 499) which was growing on 
a minute shell. In both cases it is possible to imagine that the thickenings may be 
in some way correlated with the peculiar substratum upon which the specimens are 
growing, for either on a small, readily tossed shell, or on the ever-moving appendages of 
a Crustacean, hydrorhizal tubes would be submitted to a great amount of buffeting and 
rough usage. Reaction to such abnormal external factors might result in abnormal 





Fig. 7. — Plumularia lagenifera, var. septifera. (a) Portion of stem and hydroclade. x 100. 
arising from complicated hydrorhizal growth with scattered nematothecse. 



(6) Base of stem 



strengthening of the walls. Should such a supposition be well founded, the thickenings 
in the hydrorhizal tubes could have little systematic value. Frequent nematophores, 
similar in structure to those on the remainder of the colony, arise from the hydrorhizal 
tubes at irregular intervals, but always near their borders. 

The differences between this form and P. lagenifera in size, in habit, in the intensity 
of the internal septa, in the absence of a nematophore on each stem internode, and in 
possessing chitinous thickenings in the hydrorhiza, I do not consider sufficient to warrant 
the formation of a new species. 

Measurements : — 



Stem internode, length 


0-26 mm. 


,, breadth 


0-14 „ 


Hydroclade thecate internodes, length 


0-24 „ 


„ athecate „ ,, 


0-05 „ 


Hydrotheca, depth .... 


0-066-0-090 mm 


,, diameter at margin 


0T05-0-120 „ 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 89 

Locality. — A few colonies growing on the telson and under-parts of the body of 
Palinostus lalandii (Lamk.) from Saldanha Bay, Cape Colony. 21st May 1904. 

Distribution. — Plumularia lagenifera, Allman, has been recorded from various locali- 
ties off the coast of California by Marktanner-Turneretscher (1 890, p. 255), Nutting 
(1900, p. 65), Torrey (1902, p. 77) ; from the neighbourhood of Vancouver Island by 
Allman (1885, p. 157), Nutting (I.e.) ; and from the coast of Alaska by Torrey (I.e.). 
The variety septifera has been recorded by Torrey only from Cataline Island, California. 
The general distribution of the species and its variety is thus along the shores of the 
Northern Pacific from California northwards to Alaska. That the present specimens 
should have occurred on the eastern margin of the South Atlantic is indeed remarkable ; 
but, in so mobile a group as the Hydroids, wide distribution is of little significance, and 
the above record but adds another to the long list of species which spread beyond the 
bounds of any one ocean. 

Plumularia setacea (Ellis, 1755). 

Specimens of an exceedingly minute and delicate variety of this species occur 
creeping on gulf weed in company with Aglaophenia latecarinata. They are only 7 or 8 
mm. high, and are unbranched, although in one case an appearance of bifurcation at the 
base (a phenomenon recorded by Billard in this species (1907, p. 210)) was given by 
a second colony being fixed to the first by its hydrorhiza. The hydroclade internodes, 
both thecate and athecate, are long and slender and contain two distinct septa, a distal 
and a proximal. While the rule is that a single athecate internode separates two thecate 
internodes, very rarely two intermediate internodes occur, in which case one or the other 
bears a single nematophore, the other lacking such an organ. Yet each of the internodes is 
complete as regards the internal septa, containing one at each end. They are distinct inter- 
nodes and seem to be due to spontaneous variation, for no hint could be observed that, as 
Billard found in his specimens, rupture and subsequent regeneration had taken place. 

Our specimens appear to be similar to the " distinct variety " recorded by Professor 
Nutting from gulf weed (1900, p. 57), although Nutting's specimens differ in being 
branched. 

Detailed measurements indicate that the Scotia specimens stand intermediate to the 
variety found by the Travailleur at Cape Spartel and to the south of Madeira, and the 
typical form whose dimensions Billard records. 

Measurements : — 



Height of colony 

Length of stem internodes 

Breadth of ,, „ 

Length of intermediate internodes 

,, thecate internodes . 
Breadth of ,, 



7-8 mm. 
0-33-0-39 mm. 
0-075-0-09 „ 
0-14-0-21 „ 
0-31-0-40 „ 
0-42-0-48 ., 



Locality. — Creeping on gulf weed found in spawn net at Station 538. Lat. 32° 1 1' N. , 
long. 34° 10' W. 30th June 1904. 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 4). 12 



90 



MR JAMES RITCHIE : SUPPLEMENTARY REPORT ON 



Antennopsis scotise, Ritchie, 1907 (1) . 

An additional colony of this species from the locality from which it was originally 
recorded enables me to amplify the diagnosis already given and to add a description of 
the gonosome. The new colony is of the same height as the larger of our earlier speci- 
mens, 9 cm., but it bears more branches and is altogether in better condition, although 
here also the hydroclades are in many places wanting. Branches are frequent, but are 
very irregular in position. They may bear secondary branches which are long, of uniform 
thickness, and are seldom branched. The stem and branches are composed of a thick 
bundle of tubes with transparent walls and without nodes. From short processes on the 
outermost of these the hydroclades spring, following one another on the same tube at a 





Fig. 8. — Antennopsis scotice. (a) Portion of hydroclade showing thecate and athecate internodes, and arrangement of 

nematothecse. x 65. (b) Gonotheca. x 45. 

distance of some 0'73 mm., and thus forming a close-set coat round the branches. In 
consequence of the delicacy of the hydrothecal margins a perfect hydrotheca is rare, but 
where such occurs it shows a slight widening at the mouth. 

An important addition has to be made to the previous description of the trophosome 
as regards the number of the nematophores accompanying the hydrotheca. Besides the 
median proximal nematophore and the lateral pair surmounting the processes which run 
alongside the hydrotheca, there is a supplementary lateral pair, the individuals of which, 
one on each side of the hydrotheca, rest on the upper surface of the lateral process 
almost in the corner formed between it and the internode (fig. 8). These nematophores, 
although of similar structure to, are considerably smaller than, the others, and are so 
delicate that they are frequently absent, their former position being marked only by a 
small opening in the lateral process. They correspond exactly to the supplementary 
nematophores described in Antenella quadriaurita of the present paper. 

The athecate internodes show more variation than in the former specimens, 
their length in many cases preventing the margin of a hydrotheca from reaching 
the level of the proximal end of the succeeding thecate internode. Only two 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 91 



deviations from the normal structure of the athecate internodes have been observed 
where, instead of one, two nematophores were present. 

Gonosome. — The gonangia arise laterally from the hydroclades. They are situated 
immediately beneath the hydrothecae on a short process from the hydroclade internode, 
between which and the body of the gonangium a short internode intervenes. In size 
a gonangium is three or four times as large as a hydrotheca ; in shape it is broadly 
ovate, tapering proximally into a short stalk which rests upon the intervening internode 
mentioned above, and abruptly truncated distally, where the large terminal aperture is 
closed, prior to the maturity of the contents, by a one-flapped operculum attached by 
its abcauline edge to the wall of the gonangium. Two large nematophores are present, 
one on each side of the gonangium near its base. 

Measurements : — 

0-28-0-32 mm. 



nuucw,re mucmuura, icngun 

Thecate ,, „ * . 


0-36-0-38 


Hydrotheca, length 


0-20-0-24 


„ breadth at mouth . 


0-21-0-24 


Gonangium, length 


0-88-0-91 


,, greatest breadth . 


0-53-0-57 



Locality. — As previously recorded — entrance to Saldanha Bay, Cape Colony. 25 
fathoms. 21st May 1904. 

Growing on a sponge from the same locality were several simple colonies up to 
7 mm. high. Although there are no signs of fasciculation, nor even of true branching, 
the minute structure corresponds so exactly with that of A. scotice that I cannot but 
conclude that the small colonies represent an early stage of that species. They bear no 
gonangia. 

MonostcBchas quadridens (M'Crady, 1859). 

Two minute colonies of this species were found growing on the leg of a masked 
crab. They differ from the typical form described by Nutting (1900, p. 75) only in 
their minuteness — they are less than 1 cm. high — and in the length of the athecate 
intermediate internodes of their hydroclades. Although each bears two or three hydro- 
clades, neither of the colonies is branched ; yet circular holes at the bases of some of the 
hydrothecae, indicating the point where a gonangium had been attached, show that the 
colonies are mature. The unusual position in which the colonies were growing is 
probably responsible for their smallness, for an unstable foundation is frequently 
accompanied by a dwarfed fauna. 

Measurements : — 



Stem, diameter 
Thecate internodes, length 
Intermediate internodes, length 
Hydrotheca, depth 

,, diameter at opening 



0*14 mm. 
0-42-0-46 mm. 
0-56-0-73 „ 
0-15-0-17 „ 
0-20-0-22 „ 



Length measured from straight node to furthest point of oblique node. 



92 



MR JAMES RITCHIE : SUPPLEMENTARY REPORT ON 



Locality. — Growing amongst seaweed on the leg of a masked crab. Station 81, 
Abrohlos Bank, Brazil. Lat. 18° 24' 8., long. 37° 58' W. Depth, 36 fathoms. 
Bottom doposit, coral. 20th December 1902. 



Antenella quadriaurita* sp. nov. 

A few sparse colonies which cannot be referred to any described species of Antenella 
were trawled off Gough Island. The stems, the largest of which is 14 mm. high, are 
hairlike and stand out rigidly from a creeping stolon like a group of stiff bristles. To 
the unaided eye the stem groups much resemble the figure of A. gracilis given by 
Allman (1877, pi. xxii, fig. 6), but in our specimens the stems are more delicate, and, 





Fig. 9. — Antenella quadriaurita. (a) Portion of stem, x 25. (6) Anterior aspect of a hydrotheca. 

(c) Lateral aspect of a hydrotheca. x 125. 



xl25. 



owing perhaps to mere accident, they are more irregular in size and more straggling in 
arrangement. The stem is divided into a series of alternating thecate and athecate 
internodes, the boundary lines between these being oblique and very distinct on the 
proximal side of the thecate internode, and on the distal, transverse but faintly indicated. 
On the stems examined the maximum number of hydrothecae was nine, but that a 
greater number may be borne is likely, as in both the " nine " colonies the stem was 
incomplete. The hydrothecse are borne on every other internode. As seen in profile 
they are cylindrical, having almost parallel edges, but viewed from in front they seem 
to be conical in shape, tapering rapidly to the base. A hydrotheca is rather deeper 
than broad, and for more than half its height is free from the internode. Its profile is 
straight, and its margin is very slightly everted. 

Each thecate internode bears five nematophores : one median, placed on a gently 

* Qauttuor, four ; and auritus, eared, signifying the presence of two pairs of nematothecse flanking the hydrotheca. 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 93 

raised portion of the internode on the near side of the hydrotheca ; a supracalycine pair, 
each individual of which reaches just to the margin of the hydrotheca and rests upon 
the end of an internodal projection stretching half-way across the hydrotheca ; and an 
additional pair — supplementary nematophores — placed on the upper and exterior surface 
of the supracalycine process, almost in the angle between that process and the inter- 
node. The latter are extremely fragile and readily detached, so that in many cases 
they are unrepresented but by a small pore in the wall of the supracalycine process, 
indicating where they had been attached. The intermediate internodes bear a number 
of nematophores varying from two to four, but three is the most frequent number. 
Thus, in twenty intermediate internodes examined, eight bore two nematophores, eleven 
bore three, while only one had four. On the athecate basal internodes, three of which 
generally follow one another in close succession prior to the first hydrotheca, the 
nematophores also vary. Of nine such, four had four nematophores, two had five, while 
series of two, three, and seven nematophores were represented on one internode each. 

Measurements : — 



Thecate internode, length * 
Intermediate internode, length * 
,, breadth 

Hydrotheca, depth 

,, diameter at margin 



0'46 mm. 
0-57-0-91 mm. 
0-07-0-087 „ 
0-21-0-22 „ 
0-17-0-21 „ 



This species is closely related to Plumularia secundaria (L., 1789, p. 3854), which, 
if the genus Antenella is to be preserved, and it seems a useful one from the point of 
view of practical convenience, must be transferred to that genus. The present species 
has been separated from A. secundaria on the strength of observations made by 
Marktanner-Turneretscher (1890, p. 252), Pictet and Bedot (1900, pp. 27, 28), and 
Billard (1907, p. 207), all of whom record the presence of a single nematotheca 
immediately above the hydrotheca, whereas in our specimens the hydrotheca is flanked 
by a pair of nematothecse in addition to the usual supracalycine pair. The athecate 
internodes in the Scotia species are twice the length of those in A. secundaria, and the 
other parts differ in their relative proportions, but little stress can be laid on so indefinite 
and so variable a character. I am unable to distinguish A. natalensis, Warren, 1908, 
from A. secundaria. 

Locality. — Gough Island. Trawled at a depth of 100 fathoms. Bottom deposit, 
Bryozoa and rock. 23rd April 1904. 



Aglaophenia allmani, Nutting, 1900. 

Two specimens are referred to this rare species. One is a comparatively small 
colony, 6*5 cm. high, still retaining some fragments of the hydrorhizal tubes. The 
other is a strongly fascicled branch, with a diameter of T5 mm. at its junction with 

* Measured from straight node to furthest point of oblique node. 



94 



MR JAMES RITCHIE : SUPPLEMENTARY REPORT ON 



what is probably part of the main stem, and with a height of 7'5 cm. It bears rather 
delicate alternate ramuli which leave it almost at right angles. 

While the general characters — branching, structure of hydoclades, shape and 
approximation of hydrothecse, position of nematophores — agree with the descriptions of 
Allman (1877, p. 39, pi. xxiii.,asyl. ramosa) and of Nutting, variations worthy of note 
have been observed in the last-mentioned organs. These variations seem to be mainly 
age differences. In the hydrothecse on the proximal parts of the hydroclades the mesial 
nematophore reaches to the level of the marginal teeth, and is adnate, all but the very 
tip. The supracalycine nematophores are generally cylindrical, with two apertures — 
one terminal, the other a large oval opening on that side of the upper surface which is 





Fig. 10. — Aglaophenia allmani, showing variation in hydrothecse and nematothecte. (a) Hydrothecse from the proximal 
end of a hydroclade. x 65. (b) Hydrothecse from the distal end of a hydroclade. x 65. 

towards the interior of the hydrotheca. Rarely this opening is much elongated, and 
extends through the band of perisarc which separates it from the terminal opening. In 
such a case a single continuous opening is formed and the nematophores might be 
described as "almost cylindrical." The tip of a supracalycine nematophore reaches just 
to the margin of a hydrotheca. Its axis, viewed from the side, lies at an angle of about 
45° with the stem. The nematophores, in general, agree with the type so far described. 
In those hydrothecse which occur towards the tip of the hydroclades, and which are 
therefore younger, the aperture is less oblique, while the adnate part of the hydrotheca 
is of the same length as in the older examples. As a consequence the mesial 
nematophore, although shown by measurement to be constant in length throughout the 
colony, appears to be shorter in the newly formed hydrothecse, since its tip falls con- 
siderably short of the margin. The supracalycine nematophores are of markedly greater 
length, overtopping the margin by about 0"056 mm., are quite cylindrical, and lie with 
their axis (viewed from the side) more nearly parallel to that of the internode. The 
lateral aperture is smaller and is never continuous with the terminal one, and the 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 95 

internodal septum which in the older hydrothecse marks the base of the nematophore is 
indicated in the terminal individuals only by a minute indentation. 

I think, with Nutting, that the general structure of the trophosome of the colonies 
resembles that of Lytocarpus rather than that of Aglaophenia. 

Measurements : — 



Stem internodes, length 


0-39-0-45 mm. 


„ ,, breadth 


0-22-0-24 „ 


Hydroclade internodes, length . 


0-38-0-42 „ 


Hydrothecae, length *. 


035 mm. 


,, diameter at margin 


0-17 „ 


Mesial nematophore, length 


0-28 „ 



Supracalycine nematophores, length of longest side, - ll mm. at base of hydroclade to 
0-14-0-17 mm. at distal end of hydroclade. 

Locality.— Station 81, Abrohlos Bank, Brazil. Lat. 18° 14' S., long. 37° 58' W. 
Depth, 36 fathoms. Bottom deposit, coral. 20th December 1902. 

Distribution. — Only recorded from Florida Reef, in the Gulf of Mexico (by Allman), 
and from a station in the Caribbean Sea (by Nutting). 



Aglaophenia dubia, Nutting, 1900. 

Two specimens of this species, the A. gracilis of Allman's G-ulf Stream Report 
(Allman, 1877), were collected in the same locality — one simple, reaching a height of 
6 cm., the other sparsely branched and rather longer. The anterior profile of the 
hydrotheca is not so markedly concave as in Allman's figure, nor is the mesial 
nematophore quite so long relatively to the height of the hydrotheca. The hydrothecae 
have nine teeth (Nutting says "about eight"), of which one on each side lies behind 
the supracalycine nematophores and is more acute than the others, while the anterior 
tooth is usually recurved. The number of the basal nematophores differs from that 
recorded by Allman and Nutting, for not only are two present on the front of each 
stem-internode (one close to the hydroclade and one on the proximal portion of the 
internode), while another small one lies at the base of the hydroclade, close to the former 
of those just mentioned — but, in addition, a large cup-shaped nematophore exists on the 
posterior aspect of internode, immediately behind the base of the hydroclade. Other- 
wise the specimens agree, point for point, with previous descriptions. 

Measurements : — 



Stem internodes, length 


0-52 mm 


,, ,, breadth 


031 „ 


Hydroclade internodes, length 


0-36 „ 


,, „ breadth at middle 


0-05 „ 


HydrothecBe, height 


0-32 „ 


,, diameter at margin 


0-14 „ 


Mesial nematophore, length 

1 of h Vf\ rnt.Vt pen. Tiipflsnrpfl firmer tlip i n f.pvn nrl p frmn tVip 


017 „ 

ha.Rp nf t.ViPi pavif.v t.n t".V>p 



96 MR JAMES RITCHIE: SUPPLEMENTARY REPORT ON 

On the branched specimen two branches, which appear to belong to the colony, since 
the) 7 ' lie in the same plane and leave the stem at the same angle as the true branches, 
were found, on microscopic examination, to be specimens of Halicornaria longicauda, 
arising from hydrorhizal tubes climbing upon the stem of the Aglaophenia colony. 

Locality.— Station 81, Abrohlos Bank, Brazil. Lat. 18° 14' S., long. 37° 58' W. 
Depth, 36 fathoms. Bottom deposit, coral. 20th December 1902. 

Aglaophenia heterodonta, Jaderholm, 1903. 

Dr Elof Jaderholm has described amongst the extra-European Hydroids in the 
Swedish Museum specimens of A. dichotoma (M. Sars), as distinct from a form with 
similar habit which he has named A. heterodonta. I now regard the specimens which 
were described in the earlier Scotia report under the name of A. dichotoma as examples 
of A. heterodonta. 

Additional material enables me to add to Jaderholm's description of the general 
habit of the colonies. His specimens were characterised by irregularly ramified stems 
bearing short, upward curling twigs. Our specimens exhibit two types. The first, pre- 
viously described (Ritchie, 1907, (1) pi. iii. fig. 2), is strictly dichotomous, although the 
branches may not develop equally in all parts of the colony. This type of branching 
is exactly similar to that of A. dichotoma. The dichotomously branched specimens 
were growing on a sponge, and are considerably taller (10 cm.) than Jaderholm's 
examples (3 '5 cm.). The habit of the second type is distinctly reminiscent of that of 
A. conferta, Kirchenpauer, 1872; that is to say, simple curved stems spring in pro- 
fusion from a hydrorhiza creeping upon an alga. There is no sign of branching. The 
largest of those colonies is only 18 mm. high, but that they are fully developed is 
shown by their sexual maturity, for several bear corbulse with male gonophores. The 
minute characters of the two types of colonies are identical, and agree with those of 
A. heterodonta. It may be, however, that this is but a synonym of A. conferta, 
the only characters which seem to separate the latter being the absence of an 
unpaired anterior reflexed tooth (which, however, appears to be present in Kirchen- 
pauer's figure); the outward, instead of the inward, direction assumed by the pair 
of teeth nearest the stem ; and the angled nature attributed to the supracalycine 
nematophores. However, it is only by examination of Kirchenpauer's type that 
such a question could be decided. 

That the branched and unbranched colonies are found on two different types of 
substratum probably indicates that they are environmental modifications ; the fixed, 
settled colony (that on the sponge) becoming luxuriant, while the drifting, unsettled 
colony (that growing on the alga) tends, as do so many alga-borne Hydroids, to remain 
dwarfed and of simple habit. 

The development of the corbula differs slightly from that of A. pluma as described 
by Allman (1871, p. 59) and Nutting (1900, p. 40), for the leaves develop less simul- 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 97 

taneously. While in Nutting's specimens all the leaves had made their appearance before 
even the primary pair had reached full development, here, when only five pairs are 
recognisable, the two first-formed pairs are already full grown ; and when six pairs are 
visible the earliest three have reached full development, the later ones being in a state 
of decreasing perfection. 

While some of the corbulse are wholly closed some remain partially open, their 
leaves, at least towards the tip, bearing nematophores on each side, and remaining 
separate from each other. In the latter case the gonophores are always male, in 
the former no gonophores remained ; but since maleness and openness go together, 
as they do in so many other cases, it seems probable that in this species we have 
an example of sexual dimorphism of the type described by Mr H. B. Torrey and 
Miss Martin (1906). 

The depth of a hydrotheca varies from 0'27 to 0'28 mm., its diameter at the mouth 
from 0*17 to 0*19 mm. ; measurements agreeing with those of Jaderholm. 

Locality. — On sponges and algae from the entrance to Saldanha Bay, Cape Colony. 
Depth, 25 fathoms. Bottom deposit, sand and kelp. 21st May 1904. 



Aglaophenia minima, Nutting, 1900. 

Two colonies (the larger 1 cm. high) are distinguished by their cylindrical, keelless 
hydrothecse with short projecting mesial nematophores as belonging to this species. In 
addition to the septal ridge, mentioned by Nutting, which traverses the hydrotheca- 
bearing internode at the level of the intrathecal ridge, another is evident opposite the 
base of the supracalycine nematophores. The opening of the hydrotheca is bordered 
by nine, instead of by eight teeth ; and in the mesial nematophore of some of the hydro- 
thecse a slight chitinous constriction, almost in line with the profile of the hydrotheca, 
is apparent. The nematophores on the stem internodes are arranged as follows : — A 
solitary long nematophore on the front and at the proximal end of the internode ; and in 
the angle between the hydroclade process and the stem, a double nematophore, with two 
diverging processes each bearing a terminal aperture. Only a single opening, however, 
connects the cavity of the double nematophore with that of the colony. The hydroclade 
process itself bears a large simple nematophore on its anterior surface, and to this appears 
to be due the bifurcated appearance mentioned by Nutting. 

No gonangia were present. 

Measurements : — 



Stem internodes, length 

,, diameter 

Hydroclade internodes, length . 
Hydrotheca, depth 

,, diameter at mouth 



0-45-0-56 mm. 
0-08-0-10 „ 

0-39 mm. 

0-34 „ 

0-14 „ 



Locality. — Growing amongst seaweed on the leg of a masked crab. Station 81, 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 4). 13 



98 



MR JAMES RITCHIE : SUPPLEMENTARY REPORT ON 



Abrohlos Bank, Brazil. Lat. 18° 24' S., long. 37° 58' W. Depth, 36 fathoms. Bottom 
deposit, coral. 20th December 1 902. 

Previously recorded from Little Cat Island, Bahamas (Nutting). 



Aglaophenia latecarinata, Allman, 1877. 

This common tropical species, identical, as shown by the researches of Professor 
Nutting (1900, p. 96), with the A. minuta of Fewkes (1881, p. 132), occurs 
among the Scotia collections in its usual habitat, creeping upon the fronds and 
bladders of Saragassum weed. The specimens, the largest of which are 13 mm. high, 
correspond with Nutting's description in all points but one. For while he mentions only 
two nematophores at the base of each hydroclade. I have observed in all cases four 
nematophore apertures, the two recorded by Nutting, and, in addition, lying immedi- 
ately distal to the internodal process from which the hydroclade projects, in the axil 
between it and the stem, a double nematophore, possessing two apertures, one directed 
to the right, the other to the left of the hydroclade process. The coenosarc of this 
nematophore connects with the general coenosarc of the colony through a single median 
perforation in the internode wall. A similar arrangement of nematophores occurred in 
the specimens examined by Bill ard (1907) and Versluys (1899). As the following 
comparative table shows, our specimens are. in all respects, somewhat larger than those 
described by Billard : — 





Scotia Specimen. 


Billakd's Talisman 
Specimen. 


Length of hydrocaulus 
Breadth of ,, 
Length of stem internodes 

,, hydrotheca 
Breadth of hydrotheca at mouth (excluding 

keel) . 
Breadth of keel .... 


6-13 mm. 
0-1 -013 mm. 
0-34 mm. 
0-31 „ 

0-15 „ 
0-03 „ 


5-6 mm. 
0-08-0-12 mm. 
0-25-0-30 „ 
0-27-0-30 „ 

0-135 mm. 

1 



Locality. — On floating Saragassum weed, caught in spawn net. 
Lat. 32° 11' N., long. 34° 10' W. 30th June 1904. 



Station 538. 



Halicornaria longicauda, Nutting, 1900. 

At first glance one of the specimens which I have referred to this species appears to 
be a fascicled colony with alternate branches ; but closer scrutiny shows that there is 
present a central axis with the remains of hydroclades, and that around this are grouped 
tubes which in their distal portions are continued as the so-called branches, and in the 
proximal appear to form the hydrorhizal tubes of the compound colony. The specimen 
really consists of a central axis upon which several distinct Halicornaria colonies happen 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 99 



to be creeping. The clustered bydrorhizal tubes of these create the resemblance to 
fascicling, but the structure is analogous rather to the rhizocaulom often exhibited by 
Lafoea dumosa than to a truly fascicled stem. The rhizoid nature of this compound 
stem is confirmed by another specimen (Aglaophenia dubia) from the same locality, 
where the two lower branches, at least so they appear to the eye, turn out to be distinct 
climbing colonies of H. longicauda. 

The largest of the colonies found by the Scotia is 1 1 cm. in height. The specific 
characters agree with those given by Professor Nutting, but in our specimens the 
hydrothecse appear to have undergone a greater degree of tilting forward than his 
figures indicate, while the lateral teeth are more strongly developed. The aperture is 
vertical and slightly constricted, with a sharp tooth projecting upwards and outwards 
on either side. The intrathecal ridge is well marked, arising near the middle of the 





Fig. 11.- — Halicornaria longicauda. Hydrothecse showing variation in the length of 
the median nematophore. x 110. 



adnate portion of the mesial nematophore and extending backwards at an angle of 45° 
with the axis of the hydroclade. Its free edge is slightly reduplicated. The supracaly- 
cine nematophores are short and stout, and when viewed from the front appear clearly 
above the hydrotheca, although they do not reach the margin of the aperture. The 
mesial nematophore varies greatly in length ; considerable differences may be seen even 
in two succeeding hydrothecse. Sometimes shorter than in the examples figured by 
Nutting, it may, on the other hand, extend far beyond the margin of the hydrotheca, 
and, curving gracefully upwards, its tip may reach the level of the upper portion of the 
rim. There are three cauline nematophores (not two, as Nutting implies) clustered 
about the stem process on which the hydrotheca rests. Two lie on the anterior surface, 
one at the distal, the other at the proximal side of the stem process, while one lies on 
the posterior aspect of the process itself. They are triangular in shape and large, 
normally with two apertures, one at each of the free angles of the triangle, but 
occasionally showing, as a variation, a third aperture, situated medianly between the 
normal two. 

The gonosome is unknown. 



100 MR JAMES RITCHIE : SUPPLEMENTARY REPORT ON 

Measurements : — 



Stem iutemod.es, length 


. 


0-29-0-35 mm 


,, breadth 




0-13-0-16 „ 


Hydroclade internodes, length . 




0-25-0-29 „ 


Hydrotheca, depth 




0-14 „ 


,, vertical diameter at 


margin 


0-13-0-14 „ 


,, horizontal ,, 


>> 


0-18 „ 



Locality.— Station 81, Abrohlos Bank, Brazil. Lat. 18° 14' S., long. 37° 58' W. 
Fathoms, 36. Bottom deposit, coral. 20th December 1902. 

Distribution. — The only record given by Nutting is from the Caribbean Sea near the 
Isthmus of Panama (lat. 9° 32' N., long. 79° 55' W. 36 fathoms). The present 
locality, also a shallow water one, is off the coast of Brazil near Porto Alegre — a consider- 
able southwards extension of the known range. 



LITERATURE. 

Mere mention is here made of the memoirs alluded to in this paper. For their titles reference must be 
made to such bibliographies as that in Hartlaub's " Die Hydroiden der magalhaenischen Region und chilen- 
ischen Kiiste " (below 1905), or those in Nutting's "Monographs of the American Hydroids." 

Alder, J., 1857. Trans. Tyneside Naturalists' Field Club, vol. iii. pp. 93-160. 
Allman, J. G., 1864. Ann. Mag. Nat. Hist. (3), vol. xiii. pp. 345-380. 

,, 1871. A Monograph of the Gymnoblastic or Tubularian Hydroids, London, Ray Society. 

,, 1877. Mem. Mus. Comp. Zool. Harvard, vol. v., No. 2. 

,, 1885. Jour. Linn. Soc. London, Zool., vol. xix., pp. 132-161. 

,, 1888. Report Scientific Results, " Challenger," Zool., vol. xxiii. 

Bale, W. M., 1894. Proc. Roy. Soc. Victoria, vol. vi. 
Betbncourt, A., 1899. "Deuxieme liste des Hydraires du Pas-de-Calais," Miscell. Biolog., Trav. Soc. Zool., 

Wimereux, vol. vii. pp. 1-13. 
Bonnevie, K., 1899. Den Norske Nordhavs Expedition, 1876-1878, No. 26. Christiania. 
Billard, A., 1906, (1). Actes Soc. Linn, de Bordeaux, vol. lxi. pp. 69-76. 

1906,(2). Expedition Antaretique Frangaise, 1903-1905, "Hydroides." Paris, 1906. 
,, 1907. Expeditions scientifiques du " Travailleur" et du " Talisman," tome viii. 

Bruce, W. S., 1894. "Antarctic Birds" in Knowledge, Sept. 1, 1894, pp. 208-210. 
Busk, Geo., 1850. Rep. Brit. Assoc. Adv. Sc, 20 Meet. (London, 1851), pp. 118-120. 
Clarke, S. F., 1879. Bull. Mus. Comp. Zool. Harvard, vol. v., No. 10, p. 239. 
Ellis, J., 1755. An Essay toward a Natural History of the Corallines, etc. London, 1755. 
Fewkes, J. W., 1881. Bull. Mus. Comp. Zool. Harvard, vol. viii., No. 7, p. 127. 

Hartlaub, C, 1900. "Revision der Sertularella-Arten," Abh. naturw. Ver. Hamburg, vol. xvi. pp. 1-143. 
1901. Zool. Juhr., vol. xiv., Syst., pp. 349-379. 
„ 1904. Resnltats du voyage du S.Y. " Belgica," Zool, " Hydroiden." Anvers, 1904. 

„ 1905. "Fauna Chilensis " — Supplement vi., Zool. Jahr., 1905. 

Hickson, S. J., and Gravely, F. H., 1907. " Hydroid Zoophytes, National Antarctic Expedition," Natural 

History, vol. iii. 
HlNOES, T.. 1861. Ann. Mag. Nat. Hist., ser. iii. vol. viii. p. 251. 

„ 1868. A History of the British Hydroid Zoophytes. London, 1868. 

Jadbrhoi.m, E., 1903. Arkiv for zool, utg. af Kyi. Svenska Vetenskapsakad., Bd. i., pp. 259-312. 



THE HYDROIDS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 101 

Jaderbolm, E., 1904. Archives de zool. exper. et generate, ser. iv. vol. iii. 

1905. Wissenschaftliche Ergebnisse der schwedischen Siidpolar- Expedition, 1901-1903, 
Bd. v., Lief. 8. Stockholm. 
Lamarck, J. P. B. A. de, 1836. Histoire natarelle des animaux sans vertebres, 2nd edit., 1836. 
Lamouroux, J. V. F., 1821. Exposition methodique des Genres de Tordre des Polypiers, etc. Paris, 1821. 

,, 1824. Zoologie du voyage autour du monde execute sur les corvettes V Urania et la 

Physicienne, etc., Paris, 1824, pp. 603-643. 
Linn^us, G, 1758. Sy sterna Naturee, 10th edit. 
„ 1789. „ „ 12th edit. 

M'Crady, J., 1859. Proc. Elliott Ng,t. Hist. Soc, vol. i., No. 1, p. 103. 

Makktanner-Turneretscher, G., 1890. Ann. des k. k. naturh. Ho/museums, Wien, vol. v. pp. 195-286. 
Norman, A. M., 1864. Ann. Mag. Nat. Hist., ser. 3, vol. xiii. pp. 82-90. 

Nutting, C. O, 1900. "American Hydroids," Part I. The Plumularidae, Spec. Bull. Smithsonian Inst., 

Washington, 1900. 
1904. Ibid., Part II. The Sertularidse. Washington, 1904. 
Ortmann, A. E., 1896. Grundziige der marinen Tiergeographie. Jena, 1896. 
Pallas, P. S., 1766. Elenchus Zoophytorum. Haag. 

Pictet, O, et Bedot, M., 1900. Resultats des Campagnes Sc, etc., Monaco, fascicule xviii. 
Ritchie, Jas., 1907 (1). "Hydroids of the Scottish National Antarctic Expedition," Trans. Roy. Soc. 

Edin., vol. xlv. pp. 519-545; also in Report on Sc. Res. of Voyage of S.Y. 
'Scotia,' vol. v. pp. 61-88, Edinburgh, 1909. 
1907 (2). Proc. Zoo. Soc. London, 1907, pp. 488-514. 
Thomson, J. A., 1904. "Note on the Gonostyles of Two Antarctic Siphonophora," Proc. Roy. PJtys. Soc. 

Edin., vol. xvi. pp. 19-22. 
Torrey, H. B., 1902. "University of California Publications," Zool., vol. i. pp. 1-104. 
Torrey, H. B., and Martin, A., 1906. Ibid., vol. iii. pp. 47-52. 
Versluys, J. J., 1899. Mem. Soc. Zool. France, 1899, vol. xiii. pp. 29-58. 
Warren, E., 1908. "On a Collection of Hydroids, mostly from the Natal Coast," Ann. Natal Government 

Mus., vol. i. part iii., pp. 269-355. 
Wright, T. S., 1858. Proc. Roy. Phys. Soc. Edin., vol. i. p. 450. 



TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 4). 14 



( 103 ) 



V. — The Meteorology of the Weddell Quadrant and Adjacent Areas. 
By Robert 0. Mossman. (With Five Plates.) 

(MS. received July 6, 1908. Read July 20, 1908. Issued separately June 9, 1909.) 

Having been struck by the evident importance of Antarctic phenomena in relation 
to the climatic condition of lower latitudes, I have endeavoured in the following 
discussion to give a complete representation of these conditions, and to trace their 
bearing on the climate of sub-Antarctic and temperate zones. It is only now that 
sufficient data are available for such a purpose. I have been singularly fortunate, 
not only in the opportunities which I have had for obtaining access to valuable data, 
but also in the possession of that " practical " knowledge which has resulted from 
my two years' residence in the Antarctic. In any discussion of this nature it is 
obviously an immense advantage to have studied in the field the conditions there 
obtaining. This I have been able to do also for the Arctic during two voyages under- 
taken for this specific purpose, thereby being enabled to appreciate the great diversity 
in the fundamental conditions affecting the meteorology of the two Polar regions. 

In such an inquiry as this, in which an attempt is made to break new ground, it 
is essential that the discussion should proceed on broad and general lines, in order that 
the fundamental principles involved may not be obscured by minor and negligible 
detail. Accordingly, I have restricted this inquiry to a discussion of the more 
prominent elements of climate. 

No portion of the Antarctic affords such an excellent field for the study of 
atmospheric conditions as that lying to the south and east of Cape Horn. During 
recent years this area has been the principal centre of South Polar exploration, no less 
than four expeditions having investigated the region lying between the meridian of 
Greenwich and 100° of west longitude. Moreover, we have in the South Orkney 
Station of the Argentine Meteorological Office the longest series of observations made 
south of 55° S., while data from lower latitudes, covering in each case many years, are 
also available for the numerous island and mainland stations of the Argentine and 
Chilian Republics. Valuable supplementary material has been furnished by the various 
summer voyages of Weddell, Ross, D'Urville, and others. 

The difficulties involved in this discussion are many, and will be readily recognised 
by those who have engaged in work of a similar nature. Among these are the 
comparative paucity of observations, the relatively great distance separating the spheres 
of work of the different expeditions, and the general want of synchronism in the periods 
covered by the observations. And yet, in spite of these difficulties and deficiencies, 
an attempt must be made to co-ordinate and co-relate the available information if an 

TRANS. ROY. SOC, EDIN, VOL. XLVII. PART I. (NO. 5). 15 



104 MR ROBERT C. MOSSMAN ON THE METEOROLOGY OF 

approximation towards the general principles of the atmospheric circulation is to be 
obtained. 

The region chosen is not only the best but the sole available area in the whole 
circumference of the globe south of 50° S. which offers a sufficient number of observing 
stations to warrant serious discussion. 

Maps are given (see Plates I. to IV.) showing the mean monthly and annual dis- 
tribution of pressure, temperature, and cloud. In their preparation I have drawn lines 
only for the regions furnishing sufficient data. I have also, as far as possible, en- 
deavoured to eliminate discordances due to the exceptional characteristics of individual 
months. With regard to Nordenskjold's observations at Snow Hill and Paulet Island 
in 1903, and those taken by Charcot at Wandel Island in 1904, I have been able to 
bring to bear on their discussion the experience derived from a residence during these 
years at the South Orkneys, which are, from a meteorological point of view, in the same 
region. In other words, I have tried to represent the conditions at these places as if 
they had been based on the same five years as were the South Orkney observations. 
In this connection ordinary methods of differentiation were not of much use, especially 
with regard to Charcot's 1904 observations, as this was a very exceptional year. It is 
for this reason that the Isotherms, Isonephs, and the Isobars do not always exactly 
correspond with the values given in the tables. Even with regard to the South 
Orkneys and South Georgia, I have allowed myself a little latitude in this direction. 
For the region of the " Belgica's " drift I have taken the mean monthly temperatures, 
calculated by Hann,* except that I am inclined to assign to February a lower mean 
temperature than that given by him. With reference to the observations taken at the 
Chilian coast stations from 1899-1905,+ the investigation is restricted, except as 
regards one or two elements such as rainfall, to the five years 1901-1905. The mean 
monthly temperature, pressure, humidity, and rainfall were for this series extracted 
from the general summary given in each volume. The mean cloud, wind force, and 
direction were computed from the tri-daily observations given " in extenso." Owing to 
the labour attached to work of this nature, an analysis of the winds was made only for 
Dungeness and Evangelists Island. For New Year's Island the data contained in the 
MS. sheets of observations were similarly computed from seven daily observations. 
As regards the Argentine Republic, much valuable information has been obtained from 
Mr Davis' work on the " Climate of the Argentine Republic," as well as from the 
elaborate data contained in the " Annals " of the Argentine Meteorological Office. In 
this connection I have to express my cordial thanks to Mr Davis for permission to 
utilise the South Orkney and South Georgia observations in so far as they bear on the 
subject under discussion. I am also indebted to Dr W. S. Bruce for the loan of his 
valuable meteorological log kept on board the "Balsena" during his voyage to the 
South Shetlands and Louis Philippe Land in the summer of 1892-1893. 

* Antarctic Manual, p. 44. 

t Annuario del Servicio Meteorologico de la Direccion del Territorio Maritimo, Valparaiso, 1902-1906. 



THE WEDDELL QUADRANT AND ADJACENT AREAS. 105 

Geographical Position of Areas Discussed. 

Weddell Quadrant — (See Plate V.). — The region extending from 90° W. to the 
meridian of Greenwich. This area is divided into the following subsections : — 

Bellingshausen Sea. — The area to the west of Graham's Land south of 64° S., 
extending to the 100th meridian of west longitude. Data for this region are entirely 
derived from the observations taken during the drift of the " Belgica." 

Graham's Land. — Includes Graham's Land, Danco Land, Louis Philippe Land, and 
the South Shetlands, by some geographers termed the Dirk Gherritz Archipelago. 
Data comprise a year and nine months' observations at Snow Hill, seven months' 
observations at Paulet Island, a year's observations at Wandel Island, and the summer 
observations of Forster, Ross, Bruce, and others in the vicinity of the South Shet- 
lands and Louis Philippe Land. 

Weddell Sea. — The region south of 60° S. between the meridian of Greenwich and 
50° of west longitude. Data include five years' observations at the South Orkneys, 
and the summer observations, extending over several months, made by Weddell, Ross, 
Biscoe, and particularly by the Scottish National Antarctic Expedition, which initiated 
the South Orkney series. 

Biscoe Sea. — This region, from the meridian of Greenwich to 40° B., although 
outside the specific area under discussion, is sometimes referred to. Data are available 
from the Expeditions of Moore, Biscoe, and the " Valdivia " and " Challenger." 

South Atlantic. — The area discussed is that portion west of the meridian of 
Greenwich lying south of 40° S. Data from ships are very scanty for some 
portions, especially in the region 40° to 60° S., between 0° and 30° W., as well 
as from 50° to 60° S., between longitudes of 30° and 40° W. On the other 
hand, observations covering four years are available from South Georgia Island in 
lat. 54° 30' S., and covering seven years from the Falkland Islands in 51° 41' S. 

South America. — South of 50° S. is represented by observations from numerous 
stations in Chili, Tierra del Fuego, and Patagonia, including many strictly insular 
situations such as Evangelists Island at the Pacific entrance to the Straits of Magellan, 
and Staten Island, and the neighbouring New Year's Island a little north-east of Cape 
Horn. Returns from a few places on the Pacific and the Atlantic side of the continent 
in from 40° to 43° S. are also given. In addition to the land observations, there is 
a large mass of data for both oceans in the publications of the London and other 
meteorological offices. 

General Climatic Features. 

The general climatic features of the region under discussion, viz. from 40° to 
70° S., may be briefly summarised as follows. Beginning with the mainland of South 
America, we note that the two most northerly coast stations on the Pacific side, Ancud 
and Gallera, have a large rainfall, and, relatively speaking, much cloud. The rainfall 



106 MB ROBERT C. MOSSMAN ON THE METEOROLOOY OF 

shows a distinct seasonal variation, being at a maximum in winter and a minimum in 
summer, while the pressure, cloud, and humidity curves are directly the reverse of this. 
Temperature is comparatively low for the latitude. Coming 8 degrees south we 
have the station of Evangelists Island : here also there is a very considerable precipita- 
tion, but with a maximum in summer and a minimum in winter, not, however, showing 
anything like the seasonal range of the northern stations. The climate is essentially 
insular and the summer temperature very low owing to the prevalence of the cool on- 
shore north-west winds, which blow with great force. Passing now to the Atlantic side, 
we find a great diminution of rainfall, most pronounced at inland stations such as 
Sandy Point (Punta Arenas) and Ushuaia. Here the temperature shows much greater 
amplitude than at the coastal stations, and is essentially of the continental type— - 
higher in summer and lower in winter — in harmony with the conditions prevailing in 
the interior of the Argentine provinces of Santa Cruz and Chubut. Another marked 
feature of this region is the low relative humidity and the small cloud amount, due 
to the prevailing winds having been deprived of much of their moisture by the 
mountainous region lying to the west. With regard to coastal and island stations on 
the Atlantic side, a very small difference in latitude makes a very great difference in 
climatic conditions. Staten Island, New Year's Island, and Orange Bay have all a 
considerable rainfall, much cloud and humidity, with a low and equable temperature ; 
but at Dungeness, on the Atlantic side of the straits, the rainfall is very much less, the 
skies are clearer, the air drier, and the temperature varies much more than at the 
island stations. These contrasts become much sharper as we proceed northwards along 
the Atlantic coast, where the influence of the warm Brazil current is strikingly apparent. 
Passing next to the Falkland Islands, we find a strictly oceanic climate, viz. strong 
winds, equable temperature, and a moderate rainfall, which last, however, is probably 
much greater on the West Falklands than at Cape Pembroke, the station here quoted. 
A noticeable feature here is the large amount of sunshine recorded. Between the 
Falklands and South Georgia, which lies only 2 degrees to the south, but some 22 
degrees to the east, a great difference is found. Here the influence of the Antarctic 
drift makes itself felt, the mean annual temperature being 8 degrees lower than at Cape 
Pembroke or at Staten Island. However, as regards sunshine and wind force, the 
two stations are almost identical. At the South Orkneys the influence of the Antarctic 
•drift is for the first time the predominant factor affecting climate. Here, in the low 
latitude of 61° S., the mean summer temperature is below freezing point, while in the 
winter readings below - 40° have been recorded. Perhaps the most striking example 
of the effect of the Antarctic drift on the climatic conditions in low latitudes as compared 
with the North Polar regions occurs during summer. At this season the isotherm of 
34° between the long, of 10° E. and 50° W. nowhere protrudes further south than 
58° S., and falls to about 52° to the east of the meridian of Greenwich. The most 
southerly extension of this isotherm in the Arctic is in latitude 74° N., long. 16° W., 
while to the north of Spitzbergen it lies in latitude 81° N., long. 10° E. Thus on 



THE WEDDELL QUADRANT AND ADJACENT AREAS. 107 

this meridian of 10° E. the summer isotherm of 34° is located only 540 miles from the 
North Pole, but, as we have seen, quite 2300 miles from the South Pole. Such is the 
great effect of the Antarctic drift-ice in the lowering of temperature. 

With reference to the characteristics of the strictly Antarctic climate, which form the 
principal theme of this memoir, 1 may briefly summarise the conditions as follows. 
Stations such as the South Orkneys, Port Charcot, and in a less degree Snow Hill and 
the south of the Bellingshausen Sea, are strongly influenced by the ocean to the north. 
At these places great variations of temperature and weather occur, especially in winter, 
according as the influence at work is continental or oceanic. With southerly winds the 
skies are clear, the air dry, and the temperature low ; but with a change to the north, 
exactly the reverse conditions obtain. For this reason the summers are much cloudier 
than the winters, and it is then that the maximum precipitation occurs. The tempera- 
ture does not fall so low as one would expect — from — 40° to — 45° F. The maximum 
temperature depends entirely on the situation. In the open sea or in the pack it rarely 
rises in summer more than 2 or 3 degrees above the freezing point, but near the 
land it may rise to 40° F., and at stations such as the South Orkneys and Snow Hill 
(which are subject to occasional visitations of fohn) the thermometer may rise even in 
winter to from 45° to 50° ; however, these high temperatures rarely last more than a few 
hours. In the matter of sunshine, great differences are found, — Port Charcot in 1904, 
for instance, recording 32 per cent, of the possible, with 140 sunless days, while at the 
South Orkneys only 13 per cent, was registered, with 189 sunless days. 

In this connection I have prepared a map (see PL IV. fig. 5) showing the mean tem- 
perature of the sea surface in summer. In the preparation of this I have employed all the 
existing material. The bending of the isotherms to the north is clearly shown ; also 
the higher temperature of the sea near the South Shetlands and to the west of Graham's 
Land, due to the north-east winds driving the warm surface waters before them. It 
may be incidentally mentioned that, as one would expect, there is a close relation 
between the summer sea temperature and the distribution of fog. Fogs are compara- 
tively rare between 40 and 50° S., west of 30° W., but between these latitudes, from 
30° W. to the meridian of Greenwich, fog is very frequent, amounting on the average 
to 13 per cent, of the observations. The indirect effect of the southern ice is apparent 
even to the north of latitude 40°, a little to the west of the meridian of Greenwich, on 
which longitude the greatest northern extension of the polar water occurs. (See plate 
vii. of Captain Campbell-Hepworth's paper " On the Relation between Pressure 
Temperature and Air Circulation over the South Atlantic Ocean," Meteorological Office, 
London, 1905, official No. 177.) 

Having described the broad climatic features, I now proceed to deal with the 
months in detail. 

In January there are two well-defined areas of low pressure ; one to the west of 
Graham's Land in about 64° S., the other over the Weddell Sea. Round the western 
low-pressure area the winds blow spirally inwards, being N.W. in lat. 60° S., long. 70° 



108 MR ROBERT C. MOBSMAN ON THE METEOROLOGY OF 

W., and N.E. in the vicinity of and to the south of the South Shetlands. On the west 
side of Graham's Land from 63° to 64° S. lat. winds are somewhat variable, and the 
combined values from the various summer expeditions give no great preponderance of 
any one wind. It is probable that this region is alternately influenced by the above 
two low-pressure areas. The Antarctic high is considerably reduced, and is shown by 
the isobar of 29"30 inches which covers the extreme south of the Bellinghausen Sea, 
the winds being almost wholly from the north-east, east, and south-east in 70° S., 87° 
W. The strongest winds in this region are from the north-east, and the lightest from 
the west. 

As compared with the previous month, temperature has risen everywhere, the rise 
being from 2 to 3 degrees, except at the Antarctic stations, where it amounts to 1 degree 
only. The isotherms have assumed their typical summer form, bending considerably 
to the south to the west of 50° W. A noticeable feature in this connection is the high 
temperature and dry atmosphere which prevails over Tierra del Fuego and Patagonia to 
the south and west of Sandy Point. 

At most places, except the Antarctic and strictly oceanic situations, temperature is 
now at its annual maximum. The amount of cloud has considerably increased over the 
region south of 60° S., but has diminished over the mainland of South America. The 
isoneph of 95 now appears over the northern part of the Weddell Sea, or over that 
region where the ocean temperature is relatively high. 

Rainfall is now at its annual maximum over a region running along the 53 degrees 
of latitude between the meridians of 68° and 75° W., reaching a total of 12 - 82 inches 
at Evangelists Island on the Pacific side of the straits, where north-west winds blow 
with greater frequency than during any other month of the year, and, since they are on- 
shore winds, are accompanied by heavy rainfall. 

The temperature-difference between the east and west sides of the Straits of 
Magellan is also at a maximum, Dungeness being 6°*2 warmer than Evangelists Island, 
as a result of the excessive precipitation and overcast sky prevailing at the latter region. 
At Ancud, Port Gallera, and Isle Mocha rainfall is at a minimum for the year, this 
region being now well within the influence of the South Pacific high. 

In February, the most marked feature in the pressure-distribution for the month is 
the complete recession of the Antarctic high from the area under discussion. This 
fall of pressure is most pronounced in the Bellingshausen Sea, and extends, though in 
a much smaller degree, over to Graham's Land and the Weddell Sea. Low-pressure 
areas are again shown to the east and west of Graham's Land. The former low- 
pressure area probably extends a very long way to the east, with the isobar of 29 "2 
assuming a northerly trend east of the meridian of Greenwich. Observations over a 
combined period of forty-five days show a reduction equal to 0*22 inch between situations 
in 37° W. and 64° E. longitude on the parallel of 62^-. The few observations that we 
have south of the Antarctic circle show but little difference of pressure between 23° W. 
and 50° E. 



THE WEDDELL QUADRANT AND ADJACENT AREAS. 109 

The wind observations in the Biscoe Sea indicate that the Antarctic high is situated 
to the south of 66° S. in about 40° E., and that the gradients are exceedingly steep for 
southerly winds, which blow with great force. There appears to be a well-defined 
low-pressure area in about 64° to 65° S. in 40° E. 

Temperature, as compared with January, shows little change over the region south of 
60° S., being, on the whole, very slightly higher. It reaches the maximum for the year 
at the South Orkneys, and is virtually the same as that of the surrounding ocean. The 
lowest isotherm, viz. 25°, is shown over the south of the Weddell Sea, but observations 
are too few for its position to be accurately denned. Except at the Falkland Islands 
and Sandy Point, temperature has decreased 1 or 2 degrees over the area from 40° 
to 50° S. 

Cloud amount continues very large over the greater part of the Antarctic and sub- 
Antarctic regions, but over Graham's Land (where there is a marked excess of south and 
south-east winds) a comparatively small amount is recorded, this being associated with 
a relatively dry atmosphere. 

In March, pressure has risen slightly in the Bellingshausen Sea, but has fallen over 
Graham's Land and the South Orkneys, being at its annual minimum at most places 
south of 50° S., with South Georgia as the only marked exception. 

The characteristic features of pressure-distribution remain substantially the same 
as in February, but both the eastern and western lows have deepened, and as pressure 
has given way more in the south than in the north, a general steepening of gradients 
has taken place between 50° and 60° S., causing a marked increase in gale -frequency 
to the south and south-east of Cape Horn, on the Chilian coast south of 50° S., and on 
the east side of Graham's Land. 

North-west winds are now at their annual maximum at Staten Island and the 
Falkland Islands, and over a considerable part of this area it is the stormiest month 
of the year. In the Weddell Sea the barometric gradient is comparatively slight until 
about 68° S. lat., and the " Scotia " observations, comprising two March months, show 
a steady diminution of pressure to at least 72° S. It is probable that during this 
month the barometric trough in the Weddell Sea occupies its most southerly position 
of the year, being probably located in about 70° S. lat. on the meridian of 20° W. 
Easterly gales are comparatively frequent south of 67° 30' S., pointing to very steep 
gradients at times in this region. The same conditions prevail in about 71° S. and 
93° W., where these winds blow with the maximum force of the year. 

Temperature has everywhere fallen. At Wandel Island, the South Orkneys, South 
Georgia, and the Chilian coast stations, which are largely under oceanic influence, the 
fall amounts to only 1 degree, but in the Bellingshausen Sea it reaches 5 degrees, and 
at Snow Hill as much as 9 degrees. The isotherms show a very steep thermal gradient 
between the east and west of Graham's Land. It is evident that the influence of the 
land area, which is shown to be probably located to the east in 66° S. and 47° W., is 
making itself felt. 



110 MR ROBERT C. MOSSMAN ON THE METEOROLOGY OF 

Cloud amount is still very large over the Weddell Sea, but over Antarctic areas, 
where the temperature has fallen, skies are clearing. 

In April, pressure has risen everywhere except in the south of the Bellingshausen 
Sea. The most marked rise is over Graham's Land, over which there is an area of 
relatively high pressure, dividing the low-pressure systems located to the east and west 
respectively. In these low-pressure systems the strength of the wind reaches its 
annual maximum, this being especially the case in the south of the Bellingshausen Sea, 
where the prevailing easterly winds blow with great force. Indeed, atmospheric 
conditions are everywhere very disturbed, and especially so in the South Atlantic 
and the vicinity of Cape Horn. The only marked exceptions to this are to be 
found over Graham's Land and at Dungeness, where the infrequency of gales 
is remarkable. 

There is in this month a decided tendency for the winds to veer. Thus at the South 
Orkneys south-west winds are at their annual maximum, and at South Georgia, Staten 
Island, and the Falklands the prevailing north-west winds are reduced in favour of 
west and south-west. On the other hand, at Dungeness, Evangelists Island, and 
Sandy Point northerly and north-east winds, which are infrequent for these places, have 
increased, and are at their annual maximum for the two first-mentioned stations. 
Temperature is now falling rapidly, the fall being greatest at the southern continental 
stations and least at oceanic stations in the south of South America. At places such as 
Sandy Point and Ushuaia, which are not specially under oceanic influences, the fall of 
temperature is also considerable. In this month the isotherms begin to assume their 
winter shape, and the amount of cloud shows a decided diminution north and south of 
60° S. lat., where the mixture of warm and cold air currents is most apparent. At 
Ancud, Point Gallera, and Isle Mocha the increase in rainfall is considerable, the 
precipitation being four times that of January — the driest month of the year. 

In May, the outstanding feature of the isobaric chart is the reappearance of the 
Antarctic high in a pronounced form in the south of the Bellingshausen Sea and adjacent 
continental regions. The south-east wind, although here still the strongest, has greatly 
diminished in force, and this reduction is shared by winds from other directions. Areas 
of low pressure are shown m the South Pacific and over the Weddell Sea. North-east 
winds are virtually absent at Wandel Island, which appears to be more under the 
influence of the high- than of the low-pressure area. 

South of 60° S. the temperature has fallen from 6° to 8°, the fall being greater 
on the eastern than on the western side of Graham's Land. At the South American 
and Island stations the fall is from 3 to 5 degrees, and is greatest, as in the previous 
month, at places removed from the direct influence of the sea. The mean temperature 
has fallen to zero F. a little to the south-east of Snow Hill, owing to the continental 
conditions there prevailing and the northerly extension of the pack. On the other hand, 
on the Pacific side of Graham's Land, where the oceanic iufiuence is considerable, much 
higher temperatures prevail. Cloud has very generally diminished, the cloudiest zone 



THE WEDDELL QUADRANT AND ADJACENT AREAS. Ill 

being located further north than in April. The isoneph of 85 now indicates approxi- 
mately the position of the pack edge at this season. 

A curious point in regard to the cloud-distribution in the region of the Straits of 
Magellan is shown by the circumstance that while the maximum for the year is reached 
at Sandy Point, the minimum is recorded at the neighbouring stations of Evangelists 
Island and Dungeness. 

Rainfall shows a large increase at Ancud, Point Gallera, and Mocha. 

In June, pressure has risen very generally in the Antarctic and sub-Antarctic 
regions. It has, however, fallen at the Northern Chilian stations, where it is at the 
annual minimum, while at the same time rainfall, cloud, humidity, and wind-force are at 
their annual maximum. The general distribution of pressure remains substantially the 
same as in May. The Antarctic high is in the same position, but has increased in 
height : the cyclonic areas in the Weddell Sea and South Pacific are still prominent 
features of the map, but with pressure about O'lO inch higher than in May. Tempera- 
ture has fallen about 6° at the Antarctic stations, but elsewhere only half this amount, 
if we except Ushuaia, where the increased cold of radiation has produced a fall of 5°. 
The isotherms have now assumed their normal winter position, the most marked feature 
being the steepening of the thermal gradient south of 60° S. 

Owing to the extension northward of continental conditions brought about by the 
freezing of the ocean, the amount of cloud has largely diminished, the cloudiest zone 
being indicated by the isoneph of 80, embracing and to the east of Staten Island, a 
similar high value being reached off the Chilian coast in 40° S. 

Generally speaking, relative humidity is now at its annual maximum. A marked 
feature is the excessive rainfall on the Chilian coast in 40° S., where Point Gallera has 
a mean fall for the month of 19" 1 8 inches. The precipitation diminishes north and 
south of this latitude. Staten Island also has its maximum fall for the year, viz. 6 '04 
inches. At South Georgia there occurs the minimum for the year. 

In July, the Antarctic high is shown in about 70° S. and 80° W., with a low- 
pressure area to the west-north-west and another over the AVeddell Sea. # Pressure has 
fallen over the whole area to the south of 50° S., west of the meridian of 60° W. 
To the east of 60° W. it has risen most at the Falkland Islands. 

Southerly and south-west winds are at their annual maximum over large portions of 
the region comprised between latitudes 52° and 65° S. and longitudes 72° and 65° W., 
gales from these quarters being frequent to the south of Cape Horn. 

Temperature is now at the minimum for the year, except at places such as Staten 
Island, which are strongly under oceanic influence. At this season the greatest cold is 

* The long series of observations at the South Orkneys indicate a distinct winter maximum of pressure which is 
high for the latitude as compared with that recorded in winter by the " Belgica " in long. 87° W., and it would appear 
that the Antarctic anticyclone in winter is central in about 80° S. and 40° E. That the centre is there rather than 
at the Pole is connected with the continental conditions induced by the freezing of the Weddell and Biscoe Seas 
as far north as 60° S. lat., and the relatively higher temperatures that prevail on the other side of the Pole in 
Victoria Land. 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 5). 16 



112 MR ROBERT C. MOBSMAN ON THE METEOROLOGY OF 

felt at Snow Hill, where the continental effect is most pronounced ; the South Orkneys 
also come well within the sphere of this influence. The lowest isotherm, viz. that of 
— 10°, is indicated to the east of Graham's Land in about 66° S., the relative mildness 
od the same parallel 10 degrees to the west on the Pacific side being a noticeable feature. 
The thermal gradient continues very steep in sub-Antarctic and Antarctic regions. 

The cloudiest zone is indicated in from 57° to 60° S. by the isoneph of 80, the 
lowest values being 50 over the coast of Graham's Land, and 60 in Tierra del Fuego. 

On the Pacific side of the Straits of Magellan rainfall is at its annual minimum. 

In August, cyclonic conditions prevail very generally, pressure having fallen except 
over the extreme south of South America. The fall is greatest — about a tenth of an 
inch — over Graham's Land and to the west. The isobaric chart shows a well-marked 
cyclonic area over Graham's Land, and steep gradients to the south and west of Cape 
Horn. Associated with this is a decided increase in storm frequency. North-east 
gales blow with great persistence and strength to the south of the South Shetlands. A 
small area of relatively high pressure is shown in the Bellingshausen Sea, and in this 
region, west of about 85° W. and south of 65° S., winds from all quarters blow with 
average strength, thus showing no indication of steep gradients. Temperature south of 
60° S. has risen decidedly, but in strictly oceanic situations in lower latitudes the rise is 
small. The isotherms present much the same aspect as in July, but very clearly 
indicate the greater rise of temperature in the south. 

Cloud shows a marked increase in high southern latitudes, especially to the east of 
Graham's Land, this being explained by the greater condensation accompanying the 
increased prevalence and strength of warm winds from lower latitudes. In Tierra del 
Fuego, in places removed from oceanic influence, the least amount of cloud for the year 
is recorded. 

In September, the first spring month of the southern hemisphere, the changes in 
pressure are somewhat irregular. There is a noticeable fall at the South Orkneys, and 
a slight increase, extending from 55° to 65° S., between the meridians of 67° and 55° W. 
The most interesting feature of the isobaric chart is the total absence of any indication 
of the Antarctic high, there being two large low-pressure areas, one over the Weddell 
Sea, the other to the west of and partially covering Graham's Land. Gradients over 
the larger part of the area south of 50° S. are comparatively uniform, except in the 
Cape Horn region, where they are steep. 

Temperature has risen everywhere. The rise is greatest south of 60° S., amounting 
to 5° or 6°, but elsewhere (with the exception of Ushuaia and Sandy Point) it does not 
reach half this amount. The isotherms north of from 63° S. on the 80th meridian of 
W. long, to 57° S. on the 40th meridian have opened out considerably, thereby show- 
ing that the continental influence of the frozen sea is becoming restricted. The 
thermal gradient is still very steep between the Pacific and the Weddell Sea side of 
Graham's Land. 

Cloud shows a comparatively uniform distribution, but has increased in the far 



THE WEDDELL QUADRANT AND ADJACENT AREAS. 113 

south as compared with the previous month. At Sandy Point, however, it falls to the 
annual minimum. 

In October, pressure has fallen very generally west of the 55th meridian, but at 
South Georgia and the South Orkneys it has increased. The east and west low- 
pressure areas have become merged into one, whose centre is normally located over 
Graham's Land, but which, judging from the winds at Port Charcot, is probably 
frequently central to the W.N.W. 

Gradients are steep west of the 60th meridian, and gales are at a maximum on the 
South Chilian coast and Cape Horn region, a relatively large number being from 
the S.W. 

Temperature is now rising rapidly south of 60° S., the increase being most pro- 
nounced at continental stations. The rise over Tierra del Fuego and adjacent islands 
is remarkably uniform. The isotherms continue to open out and the thermal gradients 
diminish. 

At Staten Island and Dungeness calms are at their annual maximum, a condition 
offering a marked contrast to the stormy weather on the Pacific side of the straits. 

The rainfall has increased on the Pacific side of the Magellan Straits, but in lat. 
40 S. on the Chilian coast the precipitation is only about half that of the previous 
month. 

The changes shown in the isobaric chart for November are remarkable. Pressure 
has risen greatly over Graham's Land, and to a less extent in the Bellingshausen Sea, 
while it has fallen appreciably at the South Orkneys, South Georgia, and Tierra del 
Fuego. An anticyclone is in process of formation on that portion of the Antarctic 
continent to the south of the Bellingshausen Sea and far south of Graham's Land. 
North-east winds prevail on the Pacific side of Graham's Land, and east and north-east 
in the Bellingshausen Sea, where they blow with great strength. The cyclonic area 
over the Wcddell Sea in this month is probably central about the latitude of the 
Antarctic circle in 10° W., with the isobars tending to the north on proceeding east of 
the meridian. In connection with this relatively higher pressure to the north, one may 
note that during the summer voyage of the " Valdivia" the west wind system was left 
behind in November in as low a latitude as 56° S. in 20° E., while east winds prevailed 
to 65° S. between the meridians of 20° to 60° E. On the return voyage in December 
the west winds were again experienced in lat. 55° S. 

While the position of the bottom trough on this voyage may have been located a 
degree or two further north than usual, it is still highly probable that the general 
condition is as described. That a great and widespread change of pressure distribution 
is in progress is evident. Speaking broadly, the month to month changes of pressure 
in Graham's Land and Tierra del Fuego are in the same direction, but in November, as 
will be seen, they are not only the reverse of each other, but that to a very large 
degree. Pressure has changed little at the Falklands, and has fallen much more 
on the Pacific than on the Atlantic side of the straits, the result being steep 



114 MR ROBERT C. MOSSMAN ON THE METEOROLOGY OF 

gradients for north-west winds south of 50° S. on the Pacific side, with comparatively 
slight gradients and a subnormal gale-frequency on the Atlantic side. This latter 
condition extends south of Cape Horn. 

Temperature has risen uniformly at all places south of 60° S., where the increase 
amounts to 6°. Elsewhere, within the limits of the area under discussion, the rise is 
remarkably uniform, being from 3° to 4°. The isotherms present few features of 
interest, with the exception of a relatively low temperature in from 5 V to 55° S. between 
the meridians of 65° and 70° W. This appears to be associated with an increase of 
south-west winds over the above region, a concomitant of the pressure-changes now in 
progress. 

The cloudiest zone, indicated by the isoneph of 90, is now well south of the 60th 
parallel, and on the Pacific side reaches to the Antarctic circle. North and south of 
this a decrease of cloud-amount is observable, the diminution being most marked in the 
neighbourhood of Staten Island, where it is at its minimum for the year. 

In December, the most marked change in the distribution of pressure is a small but 
general rise over Antarctic and sub- Antarctic regions. Elsewhere pressure has fallen, 
but not to any appreciable extent, being most noticeable at Sandy Point, where it 
amounts to 0*08 inch. A well-defined area of relatively high pressure is shown over 
Graham's Land (where pressure is at its annual maximum), and in the south-east of 
the Bellingshausen Sea. To the east and west of this anticyclonic region low- 
pressure areas are shown. Gradients are slight everywhere, and gale-frequency has 
greatly diminished. From about 30° W. to east of the meridian of Greenwich, in from 
50° to 60° S., gales have increased, pointing to a steepening of gradients in this region, 
south of which, however, there are no data. 

Winds from quarters to the south of west are unusually frequent in the South 
Orkneys, Tierra del Fuego, Evangelists Island, and the Falklands ; while at the latter 
station, as well as at South Georgia, south-east and east winds are at their annual 
maximum, pointing to the effect of the widespread changes of pressure referred to. 

Temperature in this, the midsummer month of the southern hemisphere, has risen 
most in the Antarctic, where the powerful influence of the midnight sun is asserting 
itself. The rise is least in the South Orkneys, where cold southerly and south-westerly 
winds blow with greater frequency than in any other month of the year. 

Cloud- amount has increased over the ocean on the Pacific side of Cape Horn south- 
ward up to the limit of observation, the maximum for the year being very generally 
reached in this area. 

While the characteristic features of the various months are as I have described, yet 
from time to time abnormal conditions prevail. I have selected December 1904 for 
special study because of the unusual pressure distribution and wind-circulation. 
Normally, there is at this season an area of relatively high pressure over Graham's Land, 
with low-pressure systems to the east and west, and an increase of pressure to the north 
of 58°, with isobars running along parallels of latitude. But in the month under 



THE WEDDELL QUADRANT AND ADJACENT AREAS. 115 

consideration the isobars west of the 65th meridian (and probably farther west) lay- 
approximately north and south, with pressure as high over Graham's Land and Cape 
Horn as at the Falklands (see Plate IV. fig. 4). At the South Orkneys, pressure, 
although above the average, was 0'30 inch lower than at Charcot's Winter Quarters, 
indicating a still lower barometer to the east. The prevailing winds over a very wide 
area south of 50° S. on the Pacific and 40° on the Atlantic side were from the south or 
south-west. That these were also the prevalent winds between the South Shetlands 
and the South Orkneys is evident by the low latitude in which the Argentine corvette 
" Uruguay," on her voyage to the South Orkneys in 1904, met with the pack, this 
being encountered in lat. 58° 40', long. 50° 40' W. The vessel arrived at the South 
Orkneys on 31st December, having forced her way through 250 miles of ice. On her 
voyage to Gerlache Straits I had an opportunity of examining the pack. I perceived 
that it showed hardly a trace of having been subjected to oceanic swell, while the 
presence of unbroken ice fields as much as 30 miles in length, not far removed from the 
outer edge of the main body, in lat. 59° 5', long. 49° to 50° W., was further proof of the 
absence of the normally prevailing westerly and north-westerly winds. I have referred 
to these facts as they bear directly on the question of the prevailing winds over the 
region and the trend of the isobars. That unusually quiet weather prevailed to the 
west of 65° W. and at Cape Horn is indicated by the total absence of strong winds at 
Dungeness and Evangelists Island, while the comparative rarity of gales at Charcot's 
Winter Quarters is also worthy of notice. At the South Orkneys, however, south-west 
gales were frequent, especially after the middle of the month. On the mainland of 
South America the most pronounced feature was the persistence of a low-pressure area 
oft" the coast of Chili in about lat. 40° S. This type of pressure-distribution prevailed on 
nineteen days of the month, but on the chart for the whole month the more distinctive 
features are masked by the normal conditions prevailing during the remainder of the 
month. This is the type of pressure-distribution associated with the presence of an 
Antarctic high at all seasons, and these highs appear to pass away to the N.E. and 
over the Atlantic. In the table on the following page are given values of the 
principal climatic elements for ten stations south of 40° S., and, as far as the data 
permit, they have been compared with the normal. 

It will be observed that at Charcot's Winter Quarters temperature was about the 
normal for the region, but that at the South Orkneys and at the island and coastal 
stations on the east side of the southern extremity of the American continent it was 
markedly below the normal, the deficit being practically the same, viz. 2°. On the 
other hand, it was markedly above the normal at Ushuaia, the excess increasing in 
a N.W. direction, and reaching 27° at Evangelists Island, where rainfall, cloud, and 
wind-force were all much below the average. This result is obviously due to the 
relative infrequency of the normal cool, onshore, rain-bearing winds, and the excess 
of land winds resulting from the abnormal distribution of pressure referred to. It will 
be observed that while southerly winds were in excess at the South Orkneys, a relatively 



116 



MR ROBERT 0. MOSSMAN ON THE METEOROLOGY OF 



Large number of N.E. and E. winds, blowing with considerable force, occurred at the 
Falklands and New Year's Island. Reference to the daily weather maps of the 
Argentine Meteorological Office show that these winds were associated with the passage 
of low-pressure systems from the Argentine coast north of 50° S. An analysis of the 
C. Pembroke barometer showed that the north-west wind experienced before the passage 
of the low-pressure area, and the south-east wind which blew after the passage of the 
low-pressure area, had practically the average pressure of the month, but that with the 



December 1904. 








Mean Pressure at 32° Sea- Level and Standard Gravity ; Mean Temperat 


ure Reduced 




to Sea-Level ; Rainfall, Cloud Amount, Wind Direction and Force 






At various Stations from, 40° to 65° South. 




















Cloud 


Percentage Frequency of Wind. 




Wind Force 


or Velocity. 


Station. 


Lat. 


Long. W. 


Pressure. 


Temp. 


Rain. 


0-10. 








1 


























] 




N. 


N.E. 


E. 


S.E. 


S. 


S.W. 


W. 


N.W. 


Calm 


Miles per 
Hour. 


Estimated 
Scale 0-12. 




. , 


. , 


Ins. 


° 


ns. 
















Port Charcot . 


65 03 


63 26 


29679 


31'2 


1-21 


8-0 


2 


19 


2 


5 


37 


22 


1 


1 


11 






South Orkneys . 


60 44 


44 39 


29-374 


28-8 


0-19 


9-5 











8 


41 


29 


15 


7 





15-5 




Cape Pembroke 


51 41 


57 43 


29-695 


45-5 




8-1 


4 


7 


13 


11 


7 


27 


23 


8 







4-3 


New Year's Island . 


54 39 


64 07 


29-693 


44-8 


1-99 


8-5 


9 


16 


11 


6 


12 


22 


14 


10 





15-5 




Ushuaia . 


54 52 


6S 07 


29-707 


50-6 


2-64 


























Harbertown 


55 00 


67 00 


29-704 


51-4 


1-39 


























Dungeness 


52 24 


68 25 


29734 


50-2 


2-38 


7-5 


6 


10 


6 


2 





29 


15 


7 


25 




2-6 


Evangelists Island * 


52 24 


75 06 


? 


49-3 


6-46 


7-7 


6 





4 


1 


10 


17 


23 


18 


21 




3-0 


An end 


41 51 


73 50 


29-984 


57-9 


3-22 


5-6 


12 


1 


2 


1 


16 


2 


27 


28 


11 




30 


Point Oallera . 


40 01 


73 44 


29-969 


55-8 


3-10 


5-1 


14 








1 


52 


3 


6 


9 


15 




3-4 


* The barometer readings are of doubtful value at this station. 








December 1904. 








Departure from the Average. 








Note. — The heavy type indicates an excess, and the italic a defect. 


















Percentage Frequency of Wind. 


Wind Force. 


Station. 


Pressure. 


Temp. 


Rain. 


Cloud. 


































N. 


N.E. 


E. 


S.E. 


S. 


8.W. 


W. 


N.W. 


Calm. 


Miles per 
Hour. 


Scale 
0-12. 




Ins. 


. 


Ins. 






















Port Charcot 


•348 







1-0 
























South Orkneys 


096 


1-8 




0-3 


6 


r 


3 


1 


19 


5 


7 


h 




2-9 




Cape Pembroke . 


084 


9-1 




09 
























New Year's Island 


•249 


20 


1-82 


11 


1 


n 


10 


h 


/, 


6 


2 


16 


6 


30 




Ushuaia 


•263 


1-3 


0-69 


























Dungeness . 


196 


0"J 


1-21 


0-7 


2 


7 


4 





2 


21 


1 


/, 


13 




0-8 


Evangelists Island 




2-7 


3-82 


0-3 


5 





3 





5 


2 





16 


15 




1-2 


Ancud .... 


■019 


o-o 


0-18 


0-2 





3 





1 


8 


8 


4 


7 


7 




0-6 


Point Gallera 


•043 


o-i 


0-11 


0-3 








1 


1 


1 





1 





6 




0-2 



east wind pressure was 0'114 inch below the normal. Captain Hep worth has referred* 
to the remarkable prevalence of easterly gales in the months of October and November, 
between the meridians of 30° and 40° W., from 45° to 50° S. It is very probable that 
these gales are due to the passage of low-pressure systems which have travelled sea- 
wards in an easterly direction from the Argentine coast. 

With respect to the bearing of the South Orkney and other Antarctic observations 
on the weather of South America, it is proposed to prepare charts for each month, 



* " The Relation between Pressure Temperature and Air Circulation over the South Atlantic," Meteorological 
Office, London, 1905, Official No. 177, p. 9. 



THE WEDDELL QUADRANT AND ADJACENT AREAS. 117 

showing the conditions over the continent for the whole period of recent Antarctic 
research. In this connection the South American data will be joined up with that 
furnished by the Falklands and South Georgia. 

Mean Atmospheric Pressure Temperature and Cloud for the Year. 

The maps showing the mean annual barometric pressure, mean temperature, and 
mean amount of cloud may be said to represent the sum and substance of the 
year's weather (see figs. 1, 2, and 3, Plate IV.). 

The isobaric charts indicate in the clearest manner the relatively higher pressure 
which prevails to the south over the Bellingshausen Sea, as compared with that over 
the Weddell Sea in similar latitudes, and the wind-circulation follows the pressure- 
distribution. The Peninsula of Graham's Land thus divides two wind systems. To 
the west we have, as at Wandel Island, a marked prevalence of strong north-easterly 
winds, due, as we have seen, to the all but permanent cyclonic area to the westward. 
To the south of this is the region embraced by the " Belgica's " drift, in which the wind- 
circulation is of a distinctly monsoonal character — easterly in summer, and westerly in 
winter. Here the easterly winds are by far the strongest, indicating the steepness of 
the barometric gradient between the Polar anticyclone and the South Pacific low- 
pressure area. The anticyclone migrates with the season — facing the Pacific Ocean in 
summer, and retreating over the continent to the south of the Indian Ocean in winter. 
While these conditions represent the normal state, different seasons appear to diverge 
considerably from the average. Thus in 1903 the Weddell Sea "low pressure" was 
deepened, and lay further to the west than usual. As pressure was about the normal 
north of 50° S., a very steep gradient was set up for W.N. W. and N.W. winds. That 
this rapid fall of pressure prevailed to at least 65° S. is shown from Nordenskjold's 
Snow Hill observations. In 1904, on the contrary, the pressure gradient was less than 
half that of the previous year, as shown by the simultaneous observations at the 
Falkland Islands and the South Orkneys. Between these two stations the mean 
difference in the barometric pressure for the nine months April to December was 
0'456 in 1903, while in the corresponding period of 1904 the difference was only 
0'199 inch. The remarkable persistence of easterly winds during the winter of 
1902 at the winter quarters of the Gauss in 90° E. indicates that the Antarctic 
anticyclone lay over the frozen continent to the south and south-west of this 
station (see footnote, p. 111). 

Pending the publication of the detailed official reports of the expeditions which have 
taken part in the recent International Antarctic campaign, it would be premature to 
draw final conclusions, but a careful analysis of all the available material enables us to 
recognise four regions characterised by permanent low-pressure areas. These are 
central in the Ross Sea in lat. 70° S., and long. 170° W. ; in the Bellingshausen Sea 
in 65° S. and 90° W. ; in the Weddell Sea in 66° S. and 30° to 35° W. ; and in the 
Antarctic Ocean in 63° S., between the meridians of 80° and 105° E. It will be seen 



118 MR ROBERT C. MOSSMAN ON THE METEOROLOGY OF 

that, with the exception of this last-named low-pressure area, these cyclonic centres are 
situated in those regions where the ocean extends furthest south and where the forces 
exerted by the mixture of the rivers of cold desiccating air from the Antarctic continent 
and the humid air of the oceans are at a maximum. As bearing on this point, it may be 
noted that every occurrence of ''Glazed Frost" seen on the voyage of the " Scotia" in 
the Weddell Sea was observed in from 67° 10' to 67° 39' S., and that at the South 
Orkneys this phenomenon is at a maximum in winter, when the theatre of cyclonic 
activity is north of its summer position. 

It will be further noticed that these low-pressure regions are situated approximately 
in the longitudes occupied by the permanent southern high-pressure belts, and, like them, 
are subject to seasonal oscillations. Judging from the summer observations, the 
winds in and near the centre of these permanent low-pressure areas are not of much 
force, the ring of steep gradients being, as one would expect, where the Antarctic 
anticyclone is most developed with reference to the contiguous low-pressure system. In 
the Weddell Sea the effect of the wind-circulation is clearly shown in the direction of 
the sea surface currents. The observations of Weddell and the "Scotia "show that 
there is a strong set to the N.W. to the south of 66° S., but in from 64° to 65° S. on 
the meridian of 40° W. the set changes to the E. and E.N.E., which is the general 
direction of the surface drift in lower latitudes. Not only so, but in the region 
dividing these two currents a large number of weatherworn bergs were seen, which, 
judging from their general appearance, had been in the vicinity for a long time. This 
region may therefore not inappropriately be termed the " Weddell Sea Doldrums." 
At the South Orkneys the whole set of the pack and of the icebergs is to the E.N.E., 
and the few cases of opposite movement recorded were probably a tidal effect. 

The circumstance that there is a distinct seasonal change in the position of these 
" centres of action " is evident. Not only do the surface winds change with the 
season, but these changes are also clearly observable at the cirrus level ; thus in summer 
the prevailing direction of the upper clouds at such places as South Georgia and the 
South Orkneys is from the west and north-west, but in winter the upper currents show 
a large movement from the south and south-west. This indicates that at the cirrus 
level (which in these latitudes is at a comparatively low altitude) the low-pressure area 
to the south-east has swung more to the north. At the " Belgica's" position in 70° to 
71 ° S., long. 80° to 90° W., exactly the reverse obtains ; and it is of interest to note that 
the maximum rate of movement was from the north-east at both seasons. The whole 
question is exceedingly interesting, but it is not safe to draw conclusions pending the 
publication of the reports of the various expeditions. 

It is unnecessary to dilate at length upon the maps of mean annual temperature and 
cloud, as the factors affecting them have been fully discussed under the monthly 
headings. Attention may, however, be called to the bending of the isotherms south- 
ward west of 55° W., as a result of the prevailing north-east winds. On the other hand, 
the crowding of the isotherms on the east side of Graham's Land is very striking. Here 



THE WEDDELL QUADRANT AND ADJACENT AREAS. 119 

the thermal gradient is exceedingly steep, as shown by the synchronous observations 
made by the Swedes at Snow Hill and the adjacent station of Paulet Island, some fifty 
miles to the north. In this region the temperature falls off at the rate of about 6° to 
7" F. for every degree of latitude : the abrupt bending of the isotherms to the south 
about the meridian of 40° W. is clearly due to the presence of the extensive land mass 
indicated by Dr Bruce on his bathy metrical chart in lat. 66° S., long. 47° W. All the 
varied information which I have collected bearing on this point affords strong evidence 
of the existence of a land mass in the assigned position. In this connection I may 
incidentally refer to some conclusions I have arrived at with reference to the probable 
existence of a peninsula or archipelago east of King Edward VII. Land. If the 
observations taken by Ross in the year 1842, between 65° and 67° S., and long. 154° to 
160° W., represent the normal meteorology of this region in summer, there is strong 
evidence that King Edward VII. Land sweeps round in a N.N.E. direction, with its 
northern apex in lat. 69° S., between the longitudes of 1 50° and 155° W. To the east of 
this the land probably falls away in a gentle curve, joining Graham's Land in 71° S. 
The whole wind circulation in the Ross Sea, as well as other meteorological concurrents, 
support the conclusion that we have to do with conditions somewhat similar to those 
obtaining in the Weddell Sea area. This calls for a peninsula situated perhaps even 
further west than I have indicated. 

The map of mean annual cloud-amount shows distinctly the distribution of this 
element in its relation to land and sea. Over Graham's Land and adjacent regions 
the skies are relatively clear, but where the pack ice and open ocean meet there is a 
zone of much cloud. The increased clearness of the skies on the east as compared with 
the west side of the continent of South America is also of interest, as is the compara- 
tively small cloud-amount of inland stations in Tierra del Fuego. 

With regard to future work, it is a matter for satisfaction that there is no abate- 
ment of interest in Antarctic investigation. A British expedition is even now at work 
in Victoria Land, where a third year's observations will materially strengthen the 
normals for that locality. The same considerations apply to Graham's Land, towards 
which a French expedition, now on the eve of departure, is to direct its efforts. Several 
prospective expeditions have announced their programmes, the realisation of which is 
earnestly to be desired. Meanwhile, the energy of the Argentine Government, as 
reflected in Mr Davis' efforts to establish the South Orkney station on a permanent 
basis, remains undiminished, while further developments, which it is hoped to realise in 
the immediate future, will enable us in time to come to discuss in a manner more 
thorough than is now possible the meteorology of the Weddell Quadrant and adjacent 
areas. 

General Summary of the Tables. 

In the remarks which follow I shall endeavour, as briefly as is consistent with 
clearness, to summarise in a general manner the contents of the tables. While the 
titles give a definite indication of the elements to which they refer, further particulars 
are desirable for the sake of clearness. Tables I. to VII. give for various stations from 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 5). 17 



120 MR ROBERT C. MOSSMAN ON THE METEOROLOGY OF 

40° to 70° S. the mean monthly and annual values of barometric pressure, air tempera- 
ture, amount of cloud, rainfall, wind velocity, relative humidity, and percentage 
frequency of wind. In their preparation I have used all the reliable material, but have 
discarded returns from stations the observations from which did not inspire one with 
confidence. Thus the barometric pressure is not given for Evangelists Island, as the 
readings are obviously in error. The data from Tierra del Fuega have been carefully 
sifted, especially with reference to mean barometric pressure. Particular trouble was 
taken to obtain good barometric means for the region around Cape Horn, which is of 
great importance as a general control on the isobars. With regard to temperature 
observations, Table II., the means given are those observed, but in plotting the values 
for drawing the isotherms the values were corrected to sea-level by adding 1*0° F. for 
every 300 feet. As only two stations exceed 150 feet, these corrections were small. 
No correction was made to bring the values to the mean of the 24 hours, as, except New 
Year's Island, which was specially treated, the means of the tri- daily observations give a 
very close approximation to the mean of the 24-hourly values. As regards New Year's 
Island, the hours of observation were 7, 8, 10 a.m., and 1, 4, 7, and 10 p.m., and the 
mean of these was brought to the mean of the 24-hourly values by adopting the values 
obtained from a reduction of the thermograph records at the adjacent station of Staten 
Island. The monthly corrections, which are all minus, are as follows : — 

Jan. Feb. Mar. April. May. June. July. Aug. Sept. Oct. Nov. Dec. Year. 
09 0°8 0°7 0-6 0°5 0°3 01 0°2 0°4 0°6 0°7 0°8 0°45 

The mean monthly and annual amount of cloud is given in Table III. 

Table IV. gives the mean monthly and annual rainfall. As regards Sandy Point and 
Dungeness, the values appear to be too low. This is due in the former instance to the 
rain gauge being at some distance above the ground. Owing to the frequency of drifting 
snow at the South Orkneys, the precipitation there cannot be given, but, on a review of 
the whole existing material, I am inclined to assign a much larger precipitation to this 
region than that given in my report.* I think about 25 to 27 inches is the probable 
annual rainfall, with a maximum in summer. As Laurie Island is on the lee side of the 
South Orkney group with reference to the prevailing westerly winds, it is probable that 
the precipitation is much greater on the neighbouring Coronation Island to the west. 

Table V. gives the mean hourly velocity in miles for the first six, and the mean 
force of the wind on the Beaufort scale 0-12, for the last seven stations. 

Table VI. gives the mean relative humidity, deduced in most cases from readings of 
the dry and wet bulb thermometer, but at Wandel Island the records are from a Richard 
Hair hygrograph. 

Table VII., showing the mean monthly and annual percentage frequency of the 
winds, has involved much labour, as the data for the South Orkneys, South Georgia, 
Evangelists Island, and Dungeness were either wholly or partially unreduced. In 
every case the values have been reduced to eight points, and the numerical results 

* Scientific Results of the " Scotia" : Meteorology, p. 274. 



THE WEDDELL QUADRANT AND ADJACENT AREAS. 121 

expressed as percentages. With regard to the seasonal values, it may be pointed out 
that Summer comprises December, January, and February ; Autumn, March, April, and 
May ; Winter, June, July, and August ; and Spring, September, October, and November. 
In place of the surface winds, the observed drift of the lower clouds at the South 
Orkneys is given. The table is not wholly satisfactory, as clear weather in winter 
usually occurs with southerly winds, so that the values for south-west, south, and south- 
east are above, and northerly winds below, the values given for this season. With 
regard to Dungeness, there would appear to be a local south-west wind at the Atlantic 
entrance to the Straits of Magellan, as this wind is unusually frequent. 

Table VIII. gives an analysis of the movement of the upper clouds (cirrus, cirro- 
cumulus, and cirro-stratus) for the months and the seasons at the South Orkneys and 
Cape Horn region, including South Georgia. For the latter station and for the 
"Belgica" only seasonal means are given. The mean direction of the upper clouds 
has been obtained by the application of Lambert's formula to the annual and seasonal 
values. The data for the South Orkneys are for the five years 1903-1907. The table 
for the Cape Horn region is from one year's observations, 1882-83, at Orange Bay and 
Ushuaia, and two years' observations at New Year's Island. Another table is given in 
which South Georgia (where twenty months' observations are available) is included 
with the above-mentioned Cape Horn data. For the " Belgica " an analysis has been 
made of the hourly cloud observations published in extenso in M. Dobrowolski's paper.* 
I may point out that the mean annual values given by me differ slightly from those 
given by M. Dobrowolski on page 58 of his report, but the difference is not of any 
moment. The cloud velocities given for the " Belgica " have also been deduced from an 
examination of the detailed hourly observations. 

Table IX. gives mean monthly and annual sea surface temperature at four places, 
viz. the South Orkneys, South Georgia, Orange Bay, and New Year's Island. The 
returns from this last-named station have been extracted from the MS. daily sheets of 
observations for the period to which they refer. 

Table X. gives an abstract by months of the observations made during summer 
expeditions on the east coast of Graham's Land and in the Weddell and Biscoe Seas. A 
comparative series of observations made at Staten Island for the same period as that 
covered by Dr Bruce's 1892-93 visit to Louis Philippe Land is also included. 

Table XI. contains an abstract of observations made in the months of February and 
March, so combined as to give means extending over a considerable number of days for 
definite areas. 

Table XII. contains similar means for the Ross Sea, calculated from the values given 
in "Contributions to our Knowledge of the Meteorology of the Antarctic." 

Table XIII. gives the percentage frequency and mean direction of the summer 
winds in various parts of the Weddell Quadrant and Ross Sea. The mean direction has 
been obtained from Lambert's formula. 

* Expedition Antarctique Beige: Resultats du Voyage du S.Y. "Belgica" en 1897-1898-1899 : " Met^orologie : 
Observations des Nuages," par A. Dobrowolski, Anvers, 1903. 



122 MR ROBERT C. MOSSMAN ON THE METEOROLOGY OF 

Table XIV. gives the mean force of the different winds for the same region on 
the Beaufort scale of 0-12. 

With regard to these Tables X. to XIV., in which the data utilised are based on a 
different number of daily observations, the values of the various expeditions have been 
weighted so as to allow for this. Thus on the "Challenger" and "Scotia" hourly 
observations were taken, while in Ross's Expedition and others the available data refer 
to observations made every four hours, and in some cases at even longer intervals. With 
regard to Biscoe's observations a numerical value (on the Beaufort scale) has been 
applied to his descriptive terms, such as "strong breeze," "moderate breeze," "gale," 
etc.; the directions given when "magnetic" have been referred to the "true" 
meridian by applying the variation for the epoch and locality of observation. 

Table XV., showing the force of the different winds during the drift of the " Belgica," 
has been prepared from the observations given in extenso in M. Arctowski's report,* 
the hours utilised being 4 a.m., 8 a.m., Noon, 4 p.m., 8 p.m., Mdt. 

Table XVI. gives an analysis of gale frequency at Evangelists Island and Dungeness, 
deduced from a detailed examination of the tri-daily observations published in extenso in 
the annual volumes of the Maritime Department of the Republic of Chili, 1901-1905. 

Table XVII. gives a condensed abstract of the mean barometric pressure, tempera- 
ture, etc. of the different winds at Cape Pembroke, Falkland Islands, and at the South 
Orkneys, for the seven months April to October 1903. 

Table XVI II. contains similar means based on the sixty-four days' observations taken 
by Dr Bruce in the vicinity of Louis Philippe Land and the South Shetlands from 
December 1892 to February 1893. 

Table XIX. shows the departure from the mean of average values of pressure, 
temperature, etc. during the cruises of the "Scotia" in the Weddell Sea during the 
summers of 1903 and 1904, for each Wind Direction. 

Table XX. gives the barometric and thermal windrose for the seasons and the year 
during the " Belgica's" drift, and the seasonal and annual departures from the normal. 
This table is based on those given by M. Arctowski on pages 36 and 38 of his work, 
but the directions have been reduced to eight points. With reference to these tables 
of windroses (Tables XVII. to XX.), it may be observed that the values have been 
reduced to eight directions by halving the intermediate points between the octants. 
The annual and seasonal means of the climatic elements are not the averages of the 
monthly values, each of which is derived from a different number of observations, but 
are the arithmetical means obtained by dividing the gross totals by the number of 
observations of the wind in question. 

Table XXI. gives the departure from the mean of the temperature of the different 
winds at Staten Island and Rawson for January, April, July, and October, and the 
mean of these four months as representing the year. The data are taken from The 
Climate of the Argentine Republic, by W. 0. Davis, p. 28. 

* Expedition Antarctique Beige : Ke'sultatsdu Voyage du S.Y. " Belgica" : " Met^orologie," par Henryk Arctowski, 
Anvere, 1904, pp. 58 to 83. 



THE WEDDELL QUADRANT AND ADJACENT AREAS. 



123 



Table I. 
Mean Monthly and Annual Barometric Pressure at 32° Sea-Level and Standard Gravity at Lat. 45°. 



20 -000 + inches. 



Station. 


Lat. 

S. 


Long. 
W. 


No. of 
Years. 


Years 
specified. 


Hours of 
Observation. 


Jan. 


Feb. 


Mar. 


Apr. 


May. 


June. 


July. 


Aug. 


Sept. 


Oct. 


Nov. 


Dec. 


Year. 




. , 


, 








Ins. 


Ins. 


Ins. 


Ins. 


Ins. 


Ins. 


Ins. 


InB. 


Ins. 


Ins. 


Ins. 


Ins. 


Ins. 


"Belgica" 


70 35 


87 22 


1 


1898-99 


hourly 


9-417 


8-96L 


9-153 


9-004 


9 426 


9 550 


9-450 


9-325 


9-298 


9-293 


9-357 


9-450 


29 307 


Port Charcot . 


65 03 


63 26 


1 


1904-05 




9-250 


9-156 


8-978 


9-476 


9-553 


9-389 


9-506 


9 329 


9-350 


9-344 


9-492 


9-679 


29-375 


Snow Hill 


64 22 


57 00 


11 


1902-03 




9-375 


9-143 


9-198 


9-280 


9-242 


9-330 


9-180 


9141 


9-242 


9-100 


9-379 


9-351 


29-247 


The two above* 


64 46 


60 13 


H 


1902-05 




9-265 


9-195 


9-125 


9-346 


9-345 


9-349 


9-289 


9 204 


9-278 


9-181 


9-424 


9-458 


29-288 


South Orkneys 


60 44 


44 39 


5 


1903-07 




9-251 


9-204 


9-157 


9-273 


9-239 


9-348 


9 395 


9411 


9-297 


9-322 


9-218 


9-288 


29"283 


South j 
Georgia i 


54 31 
54 13 


36 00 
36 33 


w 


1882-83 
1905-0S 


1882-83 „ \ 
1905-07 8-2-8/ 


9-322 


9-366 


9-371 


9-373 ] 9-352 


9-453 


9-490 


9-536 


9-531 


9-413 


9-293 


9-322 


29-402 


Orange Bay . 


55 31 


68 05 


1 


1882-83 
1875-77 | 
1882-83 \ 
1903-05 1 
1901-05 


hourly 


9-441 


9-559 


9-232 


9-455 1 9578 


9-515 


9-455 


9-417 


9-523 


9-350 


9-279 


9-464 


29-447 


Port Stanley . 


5141 


57 42 7 ■! 


4 hourly 


9-552 


9-541 


9-508 


9-624 | 9631 


9-616 


9-699 


9 655 


9-665 


9 650 


9-670 


9-610 


29-618 


Dungeness 


52 24 


1 
68 25 5 


8-2-8 


9-566 


9-502 


9-U88 


9-590 9-594 


9-511 


9-577 


9-557 


9-550 


9-561 


9531 


9-538 


29-547 


Sandy Point . 


53 10 


70 54 ( 14 


1889-1902 


7-2-9 


9-487 


9470 


9-531 


9-546 9-584 


9-680 


9-566 


9-677 


9 684 


9-619 


9-527 


9-1,1,6 


29-568 


Staten Island ) 
(Jshuaia J 


54 3S 


66 00 


18-19 


1876-1906 


7-2-9 


9-431 


9-450 


9-1,17 


9-501 9-513 


9-548 


9-521 


9-557 


9-621 


9-539 


9-439 


9-444 


29-498 


Rawson 1 

(Chabut) J 

Aucud (Chili) . 


43 1? 


65 05 


14 { 
6 


1880-88 
1896-1900 


\ 7-2-9 


9-768 


9-785 


9-809 


9-S52 


9-848 


9-959 


9-903 


9-951 


9 974 


9-892 


9-774 


9-750 


29-853 


4151 


73 50 


1900-05 


8-2-9 


9-987 


10-000 


9-972 


9-961 


9-902 


9-81,0 


9-878 


9-987 


9-996 


10 026 


10-007 


10003 


29-963 


Point Gallera . 


40 00 


73 44 


6 


1900-05 


8-2-9 


10-001 


9-992 


9-978 


9-988 


9-925 


9-869 


9-908 


10-031 


10-033 


10 062 


10-028 


10012 


29-986 



* These values are the mean of Port Charcot and Snow Hill combined with the observations made in this region by ROSS in December and January 1842- 
43, and those made by Dr Bruce and Dr Donald from December 1892 to February 1893. 

Table II. 
Mean Monthly and Annual Temperature. 



Station. 



"Belgica' 
Port Charcot . 
Snow Hill 
South Orkneys 

South Georgia 

Orange Bay . 

Port Stanley . 

Cape Pembroke 
The two above 

Ushuaia . 

staten Island 

New Year's Island 

Harhertowii . 
Sand; Point . 
Dungeness 
Evangelists Island 
Ancud 
Point Gallera 

Rawson . 

Paulet Island * 



No. of 
Years. 



1 
1 

1| 
B 

* { 
1 

*i { 

2 

»{ 



4 
15 
5 
5 
6 
7 

12 



J 
1 

7 months 



Years 
specified. 



B3 | 

37/ 



1904-05 
1902-03 
1903-08 
18S2-83 
1905-08 
1882-83 
1881-83 
1885-87 
1903-04 
As above 
1876-79 ) 
1881-83 [■ 
1904-06 ) 
1886-96 1 
1902 / 

1904-07 i 

1903-06 
1S88-1902 
1901-05 
1901-05 
1900-05 
1899-1905 
1880-88 
1898-1900 
1903 



Hours of 
Observation. 



hourly 



hourly 1882-83 

8-2-8 1905-08 

hourly 

max. andmin. 

4 hourly 
as above 



7-2-9 

7-8-10 A.M. 
1-4-7-10 P.M. 
7-2-9 
7-2-9 
8-2-9 
8-2-9 
8-2-9 
8-2-9 



Lat. 



7-2-9 ' 
3 hourly 



70 35 
05 03 
04 22 
60 44 
54 31 

54 13 

55 31 

5141 

5141 
5141 

54 52 

54 23 

54 39 

55 00 
53 10 
52 24 
52 24 
4151 
40 01 

43 17 

63 30 



Long. 
W. 


Height, 
Feet. 


Jan. 


Feb. 


Mar. 


Apr. 


May. 


June. 


July. 


Aug. 


Sept. 


Oct. 


Nov. 


Dec. 


Year. 


87 22 
63 26 
57 00 
44 39 


15 
15 
39 
20 


29-8 
32-9 
30-4 
32 1 


300 

31-2 
25-7 

322 


16-0 
29-8 
13-2 
311 


10-7 

22 6 

7-2 

25-9 


20-2 
133 

-0-8 
17-7 


4-1 
11-8 
-34 

11-7 


-10-6 

- 2-6 

- 5-5 
9-1, 


11-6 

20-4 
-30 
14-5 


-1-5 
25-7 
4-0 
19-0 


17-9 
18-6 
14-8 
24-1 


19-6 
31-5 
17-4 
29-6 


27-9 
31-2 

28-4 
30-6 


14-6 
22-2 
10-7 
23-2 


36 00 
36 33 


39 


41-8 


42-0 


39-8 


354 


32-5 


29-5 


28-2 


29-9 


32-6 


34-6 


377 


39-7 


35-3 


68 05 


46-0 


48 


42-6 


40-8 


399 


36-1 


37-8 


37 4 


(42-4) 


42-8 


44-2 


46-2 


41-7 


57 42 


22 


48-9 


49-5 


45-8 


41-4 


39-8 


36-0 


366 


37-2 


38-6 


43 


46-2 


48-0 


42-6 


57 43 
57 43 


52 
32 


48-6 

48-8 


48'8 
49-3 


46 6 
46-1 


43-9 

42-2 


39-8 
39-8 


39-2 
87-1 


35-U 
36-2 


37 6 
37-3 


39-2 
38-8 


40-6 
42-2 


44-8 
45-7 


47-0 

47-7 


42-6 
42-6 


68 07 


98 


52-8 


50-2 


46-1 


42-1 


37-3 


321 


31-7 


35-5 


40-1 


43-9 


47 9 


49-0 


42-4 


63 47 


39 


48-7 


48-0 


43-9 


43-6 


40-3 


36-9 


36S 


36-5 


38-5 


41-4 


44-6 


46-7 


421 


64 07 


164 


47-2 


46-3 


453 


41-5 


36-2 


34-8 


S/,-6 


36 6 


339 


41-4 


434 


44-4 


40-9 


67 00 
70 54 

68 25 
75 06 
73 50 
73 44 


17 

69 

10 

174 

157 

125 


52-3 
53 4 
54-2 
48 
59-5 
58-0 


49 8 
530 
52-4 
47-0 
57-6 
566 


45-9 
49-4 
49 8 
46-4 
56-0 
55-9 


426 
44-6 
466 
45-4 
53-0 
54-2 


365 
40-0 
41-2 
42-2 
50-6 
52-7 


33-9 

368 
37 1 
39-8 
47-5 

50-4 


33-0 

3U-2 
35-0 
37-8 
U5-U 
48-1 


36-4 
37-0 
36-8 
39-2 
46-4 
U8-0 


40-4 
40-6 
39-2 
40-0 
48-6 
49-8 


44-5 
445 
43-6 
41-4 
50-1 
50-3 


48-2 
48-2 
47-8 
435 
53 3 
53 


49-4 
51-2 
51-1 
461 
568 
55-3 


42-7 
44'4 
44-6 
431 
520 
52-7 


65 05 


26 


69 9 


68-2 


63-5 


641 


47-7 


hl-k 


43-0 


444 


60-9 


57-4 


62-8 


67-0 


55-9 


56 00 


30 








11-3 


3-8 


3-1 


6-9 


8-1 


7-0 


25-4 









* The means for this station are only approximate, having been measured off a curve accompanying Gosta Bodman's paper, " Om Klimatet i Antarktis 
meol siirskild hansyn till Graham Land." 

Table III. 
Mean Monthly and Annual Amount of Cloud to 10. 



Station. 



"Belgica" . 
Port Charcot 

Snow Hill . 
South Orkneys 

South Georgia i 
Orange Bay . 
Cape Pembroke . 
Sandy Point. 
Staten Island 
Evangelists Island 
Dungeness . 
Ushuaia 

Rawson . 

Ancud . 
Point Gallera 



Lat. 


Long. 


No. of 


S. 


W. 


Years. 


70 35 


87 22 


1 


65 03 


63 26 


1 


64 22 


57 00 


n 


60 44 


44 39 


5 


54 31 


36 00 


}* 


54 14 


3G33 


55 31 


68 05 


1 


5141 


57 43 


2 


53 10 


70 54 


15 


54 23 


63 47 


9 


52 24 


75 06 


5 


52 24 


68 25 


5 


54 52 


68 07 


10 


43 17 


65 05 


12 


4151 


73 50 


5 


40 00 


73 44 


5 



Years 
specified. 



1898-99 

1904-05 

1902-03 

1903-07 
/ 1882-83 
\ 1905-08 

1882-83 

1903-04 

1888-1902 

1887-95 

1901-05 

1901-05 

1876-85 
f 1880-88 I 
I 1898-1900 / 

1901-05 

1901-05 



Hours of 
Observa- 
tion. 



hourly 
2 ,, 

2 ,, 



8-2-8 
hourly 
I ., 
7-2-9 
7-2-9 
8-2-9 
8-2-9 
7-2-9 

7-2-9 

8-2-9 
8-2-9 



Jan. 


Feb. 


Mar. 


Apr. 


May. 


8-9 


9-4 


6-4 


7-2 


8-2 


9-2 


8-3 


7-9 


6-8 


5-9 


8-8 


7-6 


8-6 


6-7 


6-6 


94 


9-2 


9-2 


8-6 


8-0 


7-3 


6-7 


7-0 


7-0 


7-0 


8-6 


7-9 


8-4 


8-1 


7-6 


7-2 


6-6 


6-6 


7-2 


7-2 


6-6 


6-2 


61 


6-5 


6-7 


7-2 


7-2 


7-5 


7-9 


7-9 


79 


7-7 


7-5 


7-5 


6-9 


6-5 


6-4 


5-7 


5-8 


5-k 


6-9 


6-5 


66 


65 


6-3 


4-2 


3-8 


4-3 


4-8 


4-7 


5-1 


5-7 


6-0 


6-2 


7-6 


4-7 


h-6 


4.9 


62 


7-4 



June. 



6-5 
6-3 
6-1 
7-4 

6-8 

7-7 
7-6 
5-9 
81 
7-3 
60 
6-2 

5-0 

7-7 
81 



July. 



1,9 
5-9 
6-2 
7-3 

6-4 

7-2 



5-6 
7-5 
7-0 
5-6 
5-8 

4-9 
7-0 
7-7 



Aug. 



7-3 
8-3 
5-6 
7-6 



7-4 
7-2 
5-8 
7-5 
7-3 
5-7 
5-4 
51 

7-4 
6-9 



7-1 
7-1 



Oct. 


Nov. 


Dec. 


Year. 


8-0 


8-0 


7-9 


7-4 


7-4 


8-7 


8-0 


75 


73 


8-4 


7-6 


7-2 


8-8 


9-1 


9-2 


85 


73 


7-2 


6-8 


6-8 


7-7 


8-5 


84 


7-9 


6-6 


6-6 


7-6 


7-1 


6-6 


6-3 


6-1 


6-1 


7-0 


7-0 


7-4 


7-5 


76 


7-9 


80 


7-5 


5-6 


6-8 


6-8 


6-0 


6-5 


6-9 


7 4 


6'4 


4-4 


4"2 


4-4 


4-5 


73 


5-9 


5-8 


66 


6-1 


6-2 


54 


63 



124 



MR ROBERT C. MOSSMAN ON THE METEOROLOGY OF 



Table IV. 
Mean Monthly and Annual Rainfall. 



Station. 


Lat. 

S. 


Long. 
W. 


No. of 
Years. 


Years 
specified. 


Jan. 


Feb. 


Mar. 


April. 


May. 


June. 


July. 


Aug. 


Sept. 


Oct. 


Nov. 


Dec. 


Year. 


Port Charcot . 


65 03 


63 26 


1 


1904-06 


1-81 


1-28 


1-72 


0-93 


0-63 


1-09 


0-hl 


1-54 


1-06 


0-85 


2-29 


1-21 


14-82 


South Georgia 


(54 31 
154 14 


36 00 
36 33 


i »» 


(1882-83) 
"(1905-08) 


3 04 


5 39 


4 83 


336 


2-77 


2-07 


246 


2-26 


4-19 


3-03 


317 


3-93 


40-50 


Port Stanley . 


51 41 


57 42 


3 


1875-? 


1-97 


272 


2-17 


1-69 


2-01 


1-69 


1-46 


1-85 


1-18 


VIA 


1-34 


VIA 


20-36 


Ushuaia . 


54 52 


08 07 


10 


1876-86 


2-46 


2'46 


201 


2 73 


2-22 


2-07 


1-41 


1-46 


V26 


1-56 


1-50 


1-95 


23-09 


Staten Island 


54 23 


63 47 


10 


( 1886-93 ) 
(1899-1900) 


4-00 


5-19 


4-15 


5'25 


5-76 


604 


4-99 


3 86 


3-55 


3 -OS 


313 


4-70 


53-65 


Dungeness 


52 24 


68 25 


3 


1903-05 


2 78 


1-58 


1-47 


I'M 


0-52 


1-20 


1-18 


0-47 


0-54 


0-31 


0-49 


1-07 


12-94 


Sandy Point . 


53 10 


70 54 


15 


1888-1902 


1-57 


1-26 


156 


1-40 


1-45 


0-96 


1-25 


1-30 


1-02 


0S6 


1-15 


1-40 


15-18 


Evangelists Island 


52 24 


75 06 


7 


1899 1905 


12 82 


951 


12-10 


10-44 


7-45 


7-82 ' 0-76 


7-04 


7-67 


9-79 


9-48 


10-38 


111-26 


Ancud . 


41 51 


73 50 


6 


1900-05 


2-8U 


4"03 


6-98 


8-79 


1217 


13-24 


12-06 


10-42 


9-16 


4-79 


4-75 


3 40 


92-63 


Point Gallera 


40 01 


73 44 


7 


1899-1905 


2-92 


4-42 


6-48 


10-68 


14-72 


19 18 


1718 


12-41 


8-57 


4-77 


4-57 


3-21 


109-11 


Isle Mocha . 


38 22 


73 53 


7 


1899-1905 


0-91 


]'40 


425 


4-94 


9 01 


1139 


10-53 


6-21 


3-47 


2-00 


1-45 


1-19 


56-75 



Table V. 

Mean Houkly Velocity or Strength of the Wind. 

Instrumental Records in Miles per Hour, or Estimated Force (0-12). 



Station. 



Snow Hill 
South Orkneys 
South Georgia . 
Orange Bay 
Sandy Point . 
Staten Island . 

"Belgica" 
Cape Pembroke 
Dungeness 

Ushuaia . 

Evangelists Island 
Point Gallera . 
Ancud 



Lat. 


Long. 


No. of 


S. 


W. 


Years. 


64 22 


57 00 


If 


60 44 


44 39 


5 


54 31 


36 CO 


1 


55 31 


68 05 


1 


53 10 


70 54 


6 


54 23 


63 47 


1? 


70 35 


87 22 


1 


51 41 


57 43 


2 


52 24 


68 25 


5 


54 52 


68 07 


7 


52 24 


75 06 


5 


40 00 


73 44 


5 


41 51 


73 50 


5 



Years 
specified. 



1902-03 
1903-07 
1882-83 
1882-83 
1897-1902 
1895? 



1903-04 
1901-05 
1876-791 
1881-83) 
1901-05 
1901-05 
1901-05 



Hours of 
Observa- 
tion. 



Hourly 



4 hourly 

8-2-9 

7-2-9 

8-2-9 
8-2-9 
8-2-9 



Jan. 


Feb. 


Mar. 


Apr. 


May. 


Jun. 


July. 


Aug. 


Sept. 


Oct. 


Nov. 


Dec. 


Year. 


13-5 


157 


250 


170 


19-2 


19-2 


21-4 


19-7 


17-5 


17-0 


16-8 


.9-0 


17-6 


8-9 


142 


16-2 


16-5 


15-7 


15-7 


137 


15-0 


15-7 


16-0 


14-8 


123 


14-5 


14-0 


156 


14-9 


15-5 


14-2 


12-8 


15-7 


186 


17"2 


14-8 


1V7 


15-9 


161 


20 5 


17 5 


14-8 


13-2 


13-3 


12-4 


13-6 


13-2 


12-6 


(12-2) 


16-6 


16-8 


14-7 


9-8 


9-1 


8-2 


65 


5-8 


4-5 


6-2 


82 


8-1 


9-7 


8-5 


9-5 


7-8 


IV 5 


14-4 


19-1 


16-3 


17-0 


19-4 


19-4 


17-5 


16-6 


15-1 


13-7 


12-5 


16-0 


2-4 


2-8 


3-0 


3-5 


2-6 


24 


1-6 


2-4 


2-0 


2'7 


2-1 


2-2 


2-5 


4-5 


4-3 


4-6 


4-2 


4'4 


h-1 


u-i 


4-3 


4-4 


h-1 


46 


4-5 


4-3 


35 


4-0 


41 


3-2 


33 


3-1 


3-6 


3-9 


3-5 


3-6 


3-2 


34 


3-5 


5-6 


41 


3-8 


3-4 


34 


3-5 


3-8 


3-1 


32 


4-2 


4-3 


4 3 


3-9 


4-4 


4-4 


4-6 


4-2 


4-2 


4-4 


h-o 


4'4 


4-5 


4-5 


4-3 


4-2 


4-3 


34 


3-1 


3-0 


3-4 


37 


44 


3-7 


3-6 


33 


33 


3-4 


3-2 


3-6 


2-4 


2-4 


2-2 


2-5 


3-1 


3-5 


3'1 


3-1 


34 


3-2 


31 


2-4 


2-9 



Table VI. 
Mean Relative Humidity (Saturation = 100). 



Station. 


Lat. 

S. 


Long. 
W. 


Hours. 


No. of 
Years. 


Years 
specified. 


Jan. 


Feb. 


Mar. 


Apr. 


May. 


Jun. 


July. 


Aug. 


Sept. 


Oct. 


Nov. 


Dec. 


Year. 


Port Charcot . 


65 03 


63 26 


hourly 


1 


1904-05 


89-6 


77-1 


83-5 


86-1 


85-8 


87-8 


92-4 


89-9 


86-5 


85-9 


85-0 


85-6 


86-3 


South Orkneys 


60 44 


44 39 




2 


1903-04 


8I.-8 


86-4 


87-5 


86-0 


90 


89-5 


89-3 


87-2 


90 9 


89 3 


88-7 


87-7 


88-0 


South Georgia . 


54 31 


36 00 




1 


1882-83 


71-9 


70-9 


72-6 


77-3 




75-9 


72-6 


72-3 






76-2 


73-9 




Orange Bay 


55 31 


68 05 


.. 


1 


1882-83 


83-0 


80-6 


79-5 


83-9 


83-2 


881 


82-0 


76-1 


(75 1) 


83 3 


82-4 


83-0 


81-7 


Ushuaia . 


54 52 


68 07 


7-2-9 


5 


1876-83 


68-8 


72-4 


72-6 


74-6 


79 


841 


83-7 


81-0 


75-6 


69-7 


68-6 


72-5 


75-2 


Dungeness 


52 24 


68 26 


8-2-9 


5 


1901-05 


B9-8 


71-3 


72-9 


77-2 


79-1 


80-9 


80-6 


80-4 


78-8 


73-7 


72-6 


72-2 


76-8 


Sandy Point 


53 10 


70 54 


7-2-9 


14 


1889-1902 


70-7 


70-9 


73-1 


74-8 


798 


76-7 


810 


74-6 


81-0 


697 


67-1 


69-8 


75-5 1 


Evangelists Island . 


52 24 


75 06 


8-2-9 


2? 


1901-03 


87-1 


85-5 


8S-0 


84-0 


83-0 


85-9 


86-4 


85-6 


83-9 


R6-0 


86-3 


88-5 


85-6 


Ancud 


41 51 


73 50 


8-2-9 


5 


1901-05 


76-5 


80-5 


82 2 


84-5 


829 


862 


83-6 


84-4 


83-2 


79-7 


77-5 


77-6 


816 


Point Gallera . 


40 01 


73 44 


8-2-9 


5 


1901-05 


83-9 


84-8 


86-2 


85-6 


86-3 


86-2 


84-5 


85-0 


84-9 


81-0 


82-8 


83-8 


84-6 


Staten Island . 


54 23 


63 47 


7-2-9 


10 


1886-96 


76-6 


77-5 i 79-2 


81-6 


84-1 


86-7 


86-4 


84-1 


81-2 


78-6 


76-3 


75-6 


80-7 


Rawson . 


43 17 


65 06 


7-2-9 


12 


( 1880-88 ) 
\ 1898-1900 f 


7V6 


73-5 75-8 

1 


76-2 


73-8 


77-1 


75-5 


84-9 


78-7 


75-8 


71-6 


72-4 


76-6 




THE WEDDELL QUADRANT AND ADJACENT AREAS. 



125 



Table VII. 

Showing the Mean Monthly, Annual, and Seasonal Frequency of Wind in the Weddell Sea 

and Adjacent Regions, per Cent. 























a 




u 


















£ 


&4 

u 














J3 


h 


£1 








£ 


c 




Stations and Directions. 




3 

3 
S 
G9 
l-S 


9 


a 

a 


"S 

3 


a 

3 


9 

a 
3 

1-5 


3 


1 

3 
< 


2 
to 

p. 

<o 
m 


s 

o 


a 

> 


a 

u 

p 




si 

B 

'S 

o. 

OS 


a 
a 

3 
DO 


a 

3 
3 
< 


a 


Antartic Research Ship 


(N. 


2 


10 


*6 


7 


27 


4 


4 


8 


10 


10 


9 


2 


8 


10 


4 


13 


5 


" Belgica." 


N.E. 


28 


17 


11 


5 


18 


7 


1 


8 


16 


9 


19 


14 


13 


14 


20 


12 


5 


1 year 1898-99. 

Hourly. 

Latitudes 70° 04' S. to 71° 13' S.' 


E. 


30 


27 


16 


18 


5 


9 


8 


6 


11 


7 


20 


22 


15 


13 


26 


13 


8 


S.E. 


18 


21 


16 


21 


4 


7 


14 


6 


7 


3 


6 


16 


11 


5 


18 


11 


9 


S. 


9 


7 


15 


8 


2 


5 


14 


4 


9 


8 


5 


6 


7 


7 


7 


8 


8 


Longitudes 81° 40' W. to 

99° 27' W. 

(Mean monthly positions are 

given in Table XV.) 


s.w. 
w. 


5 
2 


4 
8 


11 
11 


6 
13 


1 
16 


14 
35 


11 
17 


11 
28 


9 

12 


16 
20 


10 
9 


14 
10 


10 
15 


12 
14 


8 
7 


8 
14 


12 
26 


N.W. 
Calm 



6 


4 
2 


4 
10 


13 

9 


20 

7 


12 

7 


7 
24 


21 

8 


8 
18 


17 
10 


5 
17 


2 
14 


10 
11 


10 
15 


2 

8 


13 
8 


14 
13 




(N. 


10 


6 


16 


11 


6 


11 


1 


8 


5 


5 


4 


2 


7 


5 


6 


11 


7 


French Antarctic Expedition. 


N.E. 


40 


12 


33 


9 


15 


19 


7 


41 


37 


22 


39 


19 


24 


33 


23 


19 


22 


1 year Feb. 1904 to Jan. 1905. 

Hourly. 

Wandel Island. 

Latitude 65° 03' S. 

Longitude 63° 26' W. 


E. 
S.E. 


10 
5 


4 
6 







1 



3 


1 
4 


1 

6 


1 
5 


9 

4 


8 
4 


13 

4 


2 

5 


4 

4 


10 

4 


6 

5 



1 


1 
5 


S. 


5 


8 


16 


40 


52 


49 


63 


29 


23 


30 


20 


37 


32 


24 


17 


36 


47 


S.W. 

w. 

N.W. 


16 
3 
3 


7 
8 
9 


9 

2 


12 

1 


12 




6 

1 


7 

6 


6 

2 


9 
1 
2 


22 
1 

2 


12 

1 


22 
1 

1 


12 

1 
2 


14 
1 
1 


15 

4 
5 


11 

1 


6 

3 




Calm 


8 


40 


24 


26 


12 


9 


9 


8 


10 


6 


7 


11 


14 


8 


19 


21 


9 


South Orkneys. 


fN. 


4 


11 


6 


2 


5 


4 


2 


3 


2 


5 


6 


6 


5 


4 


7 


4 


3 




N.E. 


3 


8 


3 


1 


4 


2 


1 








2 


5 


5 


3 


2 


5 


3 


1 


3 years 1903-05. 


E. 


6 


2 


1 


1 


2 


1 





1 





2 


7 


3 


2 


3 


4 


1 


1 


Hourly. 


S.E. 


5 


4 


2 


1 


4 


1 





3 


3 


1 


6 


7 


3 


3 


6 


2 


1 


Latitude 60° 44' S. 


S. 


8 


7 


5 


11 


10 


9 


7 


12 


5 


3 


3 


22 


8 


4 


12 


9 


9 


Longitude 44° 39' W. 


S.W. 


28 


21 


23 


40 


34 


31 


36 


31 


32 


25 


17 


24 


29 


25 


24 


32 


33 


Lower Clouds. 


w. 


39 


33 


38 


33 


22 


38 


42 


32 


46 


39 


33 


22 


35 


39 


31 


31 


37 


Calms from Surface Wind 
Observations. 


N.W. 
Calm 


7 
6 


14 
6 


22 

7 


11 
8 


19 
9 


14 
13 


12 
9 


18 
6 


12 
6 


23 

6 


23 

5 


11 
7 


15 

7 


20 
6 


11 

6 


18 
8 


15 
9 




fN. 


12 


17 


19 


30 


17 


13 


11 


14 


20 


31 


25 


19 


19 


25 


16 


22 


13 


South Georgia. 


N.E. 


7 


10 


3 


1 


1 


7 


3 


8 


8 


10 


6 


9 


6 


8 


8 


2 


6 


4 Years 1882-83. Hourly. 


E. 


9 


6 


3 


2 


1 


4 


3 


3 


3 


6 


4 


9 


4 


4 


6 


2 


3 


1905-07. 8-2-8. 


S.E. 


12 


7 


4 


6 


4 


9 


9 


8 


2 


6 


7 


13 


7 


5 


9 


5 


9 


Latitude 54° 31' S. 


S. 


4 


3 


2 


2 


6 


2 


2 


2 


4 


6 


4 


3 


3 


5 


3 


3 


2 


Longitude 36° 00' W. in 1882-83. 


S.W. 


7 


10 


6 


5 


7 


6 


7 


5 


7 


5 


8 


5 


7 


7 


8 


6 


6 


Latitude 54° 13' S. 


w. 


23 


18 


22 


19 


23 


15 


19 


21 


27 


14 


21 


16 


20 


21 


20 


21 


18 


Longitude 36° 33' W. in 1905-07. 


N.W. 
Calm 


26 


29 


41 


35 


41 


44 


46 


39 


29 


22 


25 


26 


34 


25 


30 


39 


43 




N. 


18 


15 


22 


21 


16 


28 


17 


21 


21 


18 


24 


12 


19 


21 


15 


20 


22 


Cape Pembroke Lighthouse, 


N.E. 


15 


5 


6 


11 


4 


12 


7 


9 


13 


10 


16 


10 


10 


13 


10 


7 


9 


Falkland Islands. 


E. 


1 


8 


1 


2 


1 


1 


2 


3 


3 


1 


2 


9 


3 


2 


5 


1 


2 


2 years 1903 and 1904. J 
4 Hourly. 
Latitude 51° 41' S. 
Longitude 57° 43' W. 


S.E. 
S. 


4 
5 


2 
5 


1 
2 


3 

4 


2 

8 


3 
9 


2 
6 


1 

4 




1 


3 

8 


1 

o 


6 
5 


2 
5 


1 

4 


4 
5 


2 

5 


2 
6 


S.W. 

w. 

N.W. 


18 
20 
19 


16 
21 
28 


6 
27 
35 


14 
23 

22 


14 
24 
31 


5 

18 
24 


12 
24 
30 


11 
26 

25 


11 
29 
22 


14 
22 
24 


13 
16 
26 


20 
21 

17 


13 
23 
25 


13 
22 
24 


18 
21 
22 


12 
24 
29 


10 
23 
26 




Calm 












••• 




i 










... 











* Includes 14 days of March 1899 as well as March 1898. 



126 



MR ROBERT C. MOSSMAN ON THE METEOROLOGY OF 



Table VII. — continued. 







. 


u 














C 

.a 


C 




£2 












Stations and Directions. 




u 
es 

3 
c 

es 

1-5 


es 
3 
.a 


o 
u 
es 

3 


p. 


si 


0> 

a 
s 


3 

1-5 


s 
to 

3 


S 
o 

ft 

CO 


to 

a 
o 
o 
O 


s 

> 

o 


S 

o 


u 
es 


bi) 
a 

p. 

CO 


g 

a 

3 
CO 


S 

3 
3 


c 




fN. 


11 


15 


10 


14 


11 


12 


8 


8 


12 


13 


14 


8 


11 


13 


11 


12 


9 




N.E. 


2 


2 


2 


1 


3 


5 


7 


8 


3 


2 


4 


5 


4 


3 


3 


2 


7 


Staten Island. 


E. 


2 


1 


1 





1 


3 


1 


1 


1 


1 


1 


1 


1 


1 


1 


1 


2 


10 years 1886-96. 


S.E. 


9 


3 


4 


4 


7 


7 


10 


9 


5 


6 


7 


10 


7 


6 


7 


5 


9 


Hours 7-2-9. 


S. 


16 


10 


14 


21 


17 


13 


13 


14 


16 


13 


18 


16 


15 


16 


14 


17 


13 


Latitude 54° 23' S. 


s.w. 


15 


22 


18 


16 


14 


18 


16 


14 


17 


14 


14 


16 


16 


15 


18 


16 


16 


Longitude 63° 47' W. 


w. 


15 


18 


17 


16 


20 


16 


14 


14 


15 


15 


11 


12 


15 


14 


15 


18 


15 




N.W. 


25 


23 


29 


22 


24 


20 


23 


26 


25 


26 


25 


26 


25 


25 


25 


25 


23 




Calm 


5 


6 


5 


6 


3 


6 


8 


6 


6 


10 


6 


6 


6 


7 


6 


4 


6 




r N. 


7 


11 


9 


11 


13 


10 


13 


12 


14 


8 


5 


11 


10 


9 


10 


11 


12 


Punta Arenas 


N.E. 


10 


8 


10 


15 


17 


15 


8 


14 


12 


13 


14 


11 


12 


13 


9 


14 


13 


(Sandy Point). 


E. 


5 


10 


4 


4 


6 


4 


5 


3 


3 


1 


3 


9 


5 


2 


8 


5 


4 




S.E. 


2 


4 


3 


2 


3 


3 


6 


5 


4 


3 


5 


3 


3 


4 


3 


3 


4 


6 years 1897-1902. 

Hourly. 

Latitude 53° 10' S. 

Longitude 70° 54' W. 


S. 


5 


4 


8 


4 


8 


8 


4 


8 


7 


10 


8 


7 


7 


8 


6 


6 


6 


S.W. 


18 


24 


21 


19 


17 


19 


27 


23 


24 


28 


25 


16 


22 


26 


19 


19 


23 


w. 


32 


21 


34 


28 


20 


27 


19 


21 


20 


16 


15 


18 


23 


17 


24 


27 


22 


N.W. 


21 


18 


11 


17 


16 


14 


18 


14 


16 


21 


25 


25 


18 


21 


21 


15 


16 




Calm 






































r N. 


8 


2 


4 


9 


5 


6 


4 


6 


7 


5 


8 


4 


6 


7 


5 


6 


5 




N.E. 


4 


1 


3 


4 


2 


5 


4 


4 


6 


3 


6 


3 


4 


5 


2 


3 


4 


Dungeness (Lighthouse). 


E. 


2 





1 


2 


2 


3 


5 


1 


3 


1 


1 


2 


2 


2 


1 


2 


3 


5 years 1901-05. 


S.E. 


2 


3 


2 


1 


4 


3 


5 


2 


2 





2 


2 


2 


1 


3 


2 


4 


Hours 8-2-9. 


S. 


2 


5 


3 


2 


7 


9 


6 


5 


4 


2 


2 


2 


4 


3 


3 


3 


6 


Latitude 52° 24' S. 


S.W. 


50 


56 


51 


41 


38 


32 


42 


35 


33 


42 


41 


50 


42 


38 


52 


44 


36 


Longitude 68° 25' W. 


W. 


10 


13 


15 


15 


19 


13 


16 


26 


18 


18 


15 


14 


16 


17 


12 


16 


19 




N.W. 


9 


9 


12 


12 


14 


14 


9 


8 


13 


11 


11 


11 


11 


12 


10 


13 


10 




^Calm 


13 


11 


9 


14 


9 


15 


9 


13 


14 


18 


14 


12 


13 


15 


12 


11 


13 




fN. 


8 


5 


9 


16 


5 


11 


6 


8 


8 


6 


9 


11 


9 


7 


8 


10 


8 




N.E. 








1 


3 


2 


4 


2 


1 


2 





1 





1 


1 





2 


2 


Evangelists Island (Lighthouse). 


E. 








1 


4 


4 


5 


7 


7 


3 








1 


3 


1 


1 


3 


7 


5 years 1901-05. 
Hours 8-2-9. 

Latitude 52° 24' S. 


S.E. 

S. 

S.W. 


1 

5 

13 


1 
10 
22 


2 
10 
16 


6 

8 
12 


10 
17 
15 


6 
17 
16 


8 
14 
17 


2 
10 
19 


4 
11 
13 


1 

8 

16 


1 

5 

13 


1 

5 

19 


3 

10 
16 


2 

8 

14 


1 
6 

18 


6 
12 
14 


5 

14 
17 


Longitude 75° 06' W. 


w. 


26 


27 


26 


15 


20 


12 


12 


23 


22 


26 


23 


23 


21 


24 


25 


20 


16 


N.W. 


41 


28 


31 


24 


19 


24 


22 


23 


33 


35 


40 


34 


30 


36 


35 


25 


23 | 




Calm 


6 


7 


4 


12 


8 


5 


12 


7 


4 


8 


8 


6 


7 


7 


6 


8 


8 




fN. 


11 


12 


25 


17 


24 


25 


13 


10 


12 


15 


14 


9 


15 


14 


11 


22 


16 


Ushuaia 


N.E. 


2 


4 


4 


3 


2 


3 


3 


6 


7 


4 


3 


2 


3 


4 


2 


3 


4 


Tierra del Fuego. 


E. 


4 


4 


6 


8 


10 


7 


11 


9 


15 


13 


11 


11 


9 


13 


7 


8 


9 


S.E. 


2 





4 


2 


6 


2 


5 


2 


3 


2 


4 


3 


3 


3 


2 


4 


3 


4J years 1876-83. 


S. 


1 


1 


2 


6 


4 


9 


6 


5 


4 


1 


3 


3 


3 


3 


2 


3 


7 


Hours 7-2-9. 


S.W. 


20 


17 


15 


24 


18 


21 


18 


37 


15 


16 


16 


20 


19 


16 


19 


19 


25 


Latitude 54° 52' S. 


w. 


56 


55 


35 


33 


26 


30 


36 


19 


30 


39 


41 


49 


40 


37 


53 


32 


29 


Longitude 68° 07' W. 


N.W. 
Calm 


4 


7 


9 


7 


10 


3 


8 


12 


14 


10 


8 


3 


8 


10 


4 


9 


7 

i 



THE WEDDELL QUADRANT AND ADJACENT AREAS. 



127 



Table VIII. 

Percentage Frequency of Upper Clouds. 

Cirrus, Cirro-Cum., and Cir.-Str. 







>> 


>, 


















to 

a 








u 


a 




Stations and Directions. 




B 

3 
C 
« 


u 
XI 
9 
to 

4 


JS 

6 


'C 
p. 

«: 

2 


11 


s 
s 

1-5 

7 


3 

1-5 

7 


3 
SO 

3 
■< 

4 


a 
3 


O 

■s 

o 
1 


a 

> 
o 

5 


a 

o 

p 




u 
o 

4 


bo 

a 
'B 
» 

3 


a 
s 

3 
CO 

2 


a 

3 
"3 
< 

6 


o 

c 
6 




(N. 


2 




N.E. 


3 


2 





3 


1 





4 


7 





3 


3 





2 


2 


2 


1 


4 




E. 


3 


5 





1 


6 


2 


9 


7 








2 


1 


3 





3 


1 


6 


South Orkneys, 1903-1907 . < 


S.E. 

S. 


3 

16 


1 

8 


8 
5 


2 
17 


7 
14 


7 
22 


8 

12 


2 
12 




11 




1 


1 
1 


2 
2 


3 

10 


1 
4 


2 
8 


6 
12 


6 
16 




s.w. 


12 


23 


31 


20 


15 


30 


17 


26 


32 


27 


14 


16 


22 


25 


17 


21 


22 




w. 


48 


32 


33 


32 


21 


24 


32 


20 


39 


57 


56 


63 


39 


50 


51 


29 


26 




N.W. 


13 


25 


17 


23 


25 


8 


11 


22 


15 


11 


18 


16 


17 


15 


15 


24 


14 


Number of Observations . 




99 


113 


86 


89 


98 


120 


84 


95 


118 


105 


87 


133 


1227 


310 


345 


273 


299 




fN. 


3 


11 


fi 


8 


15 


17 


33 


34 


?,h 


19 





7 


12 


14 


6 


9 


26 


Cape Horn Region. 


N.E. 


2 











5 








3 














1 





1 


1 


1 


From observations at Orange 


E. 




















9 





























3 


Bay and Ushuaia during the 
year 1882-83, and at New 


S.E. 





















































S. 


4 


8 





13 


10 


9 











8 





1 


5 


3 


4 


9 


4 


Year's Island, December 1904 


S.W. 


21 


6 





14 


22 


20 








20 


13 


18 


18 


14 


17 


16 


13 


9 


to March 1907. 


w. 


46 


56 


47 


42 


33 


43 


58 


50 


15 


39 


48 


55 


45 


35 


52 


41 


49 




N.W. 


24 


19 


47 


23 


15 


11 





13 


40 


21 


34 


19 


23 


31 


21 


27 


8 


Number of Observations . 




39 


26 


18 


31 


20 


23 


12 


16 


20 


24 


22 


37 


288 


66 


102 


69 


51 




fN. 


5 


4 


8 


6 


8 


15 


21 


14 


27 


15 


3 


5 


9 


5 


7 


16 


12 




N.E. 


4 











3 








7 


4 


6 





3 


2 


2 


1 


3 


3 


Cape Horn Region. 


E. 




















5 


1 


4 











1 








2 


1 


As above but including South •< 
Georgia, 1882-83 and 1907-08. 


S.E. 














20 

















1 





1 





6 








S. 


2 


3 





9 


28 


10 


2 


5 





6 


1 


4 


5 


3 


12 


b 


2 


S.W. 


16 


20 


13 


10 


12 


23 


11 


11 


23 


8 


12 


18 


15 


18 


12 


lb 


13 




w. 


47 


59 


39 


45 


21 


40 


50 


43 


11 


29 


39 


48 


42 


51 


35 


44 


30 




N.W. 


26 


14 


40 


30 


8 


12 


11 


19 


31 


36 


44 


22 


25 


21 


27 


15 


39 


Number of Observations . 




75 


75 


46 


44 


41 


26 


22 


44 


26 


36 


53 


65 


592 


154 


215 


131 


92 



Seasonal Percentage Frequency of Upper Clouds. 




South Georgia. 






Spring. 


Summer. 


Autumn. 


Winter. 


Year. 


N. 





1 


3 


2 


2 


N.E. 


7 


2 








2 


E. 


4 














S.E. 





1 


17 





4 


S. 





2 


20 


10 


7 


S.W. 


7 


23 


14 


30 


19 


W. 


32 


51 


31 


46 


44 


N.W. 


50 


20 


15 


12 


22 



Mean Direction of the Upper Clouds. 




"Belgica." 


South 
Orkneys. 


South 
Georgia. 


Cape Horn 

with South 

Georgia. 




, 


o 


o 


„ 


Spring . 


S. 86 W. 


S. 76 W. 


N. 47 W. 


N. 82 W. 


Summer . 


S. 10 W. 


S. 85 W. 


S. 87 W. 


N. 76 W. 


Autumn . 


S. 61 W. 


S. 86 W. 


S. 42 W. 


N. 82 W. 


Winter . 


S. 69 W. 


S. 68 W. 


S. 68 W. 


N. 85 W. 


Year 


S. 66 W. 


S. 81 W. 


N. 89 W. 


N. 77 W. 


Mean Annual Dir 


ection, Ushuaia . . . . N. 81 W. 


I 1 !» 


,, New Year's Island . . N. 74 W. 


») >» 


,, & Orange Bay N. 68 W. 





"Belgica." 




Mean Rate of Movement, Upper Clouds. Scale 0-5. 






N. 


N.E. 
2-0 


B. 
1-7 


S.E. 
2-7 


s. 
2-5 


S.W. 
2-5 


w. 

2'2 


N.W. 


Mean. 


Spring 




23 


2-0 


2-3 


Summer 




10 


2-5 


2-2 


1-5 


1-8 


1-6 


1-3 


1-6 


17 


Autumn 








1-0 


1-0 


17 


1-5 


V8 


2 3 


1-7 


Winter 






3 






1-5 


2-4 


2-3 


2-1 


23 


Year. 




i-8 


2-2 


1-9 


1-7 


2-0 


2-0 


2-1 


2-0 


2-1 



TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 5) 







" Belgica." 






Seasonal 


Percentage of Upper Clouds. 






Spring. 


Summer. 


Autumn. 


Wiuter. 


Year. 


N. 


3 


3 






2 


N.E. 


10 


3 




3 


5 


E. 


3 


7 


4 




3 


S.E. 


3 


23 4 




8 


S. 


10 


29 


20 


5 


15 


S.W. 


34 


17 


33 


51 


34 


W. 


12 


4 


21 


9.0, 


13 


N.W. 


25 


14 


18 


21 


20 



18 



128 



Ml! ROBERT C. MOBSMAN ON THE METEOROLOGY OF 



Table IX. 

Sea Surface Temperature. 











1 
Hours 




























Station. 


Lat. 
S. 


Long. 
W. 


No. of 
Years. 


Years of un- 
specified, serva- 
Hon. 


Jan. 


Feb. 


Mar. 


Apr. 


May. 


Jun. 


July. 


Aug. 


Sept. 


Oct. 


Nov. 


Dec. 


Year. 




O / 


O / 


i A.M. 


o 





o 


o 


. 


o 


o 


o 


° 


• 


• 


■ 


- 


South Orkneys . 


60 44 


44 39 


5 


1903-07 


8 


31-8 


32-2 


32-0 


30-5 


29-4 


28 '7 ; 28 '7 


28-7 


28-8 


29-0 


29-8 


31-3 


30-1 


South Georgia . 


54 31 


36 00 


1 


1882-83 


8 


39-4 


40 1 


37-5 


34-2 


32-2 


80-1 32'8 


34-2 


30-2 


33-8 


35-9 


37-8 


34-8 


Orange Bay 


55 31 


68 05 


1 


1882-83 


8 


49-7 


50-2 


46-9 


45-5 


43-0 


42-2 j 42-1 


4V0 


(41-5) 


42-4 


46-2 


48-7 


45-0 


Nt \v War's Is. . 


54 39 


64 07 


H 


1904-07 


7 or 8 


477 


46-8 


46-4 


43 9 


41-5 


39 6 \38-6 


39-5 


41-1 


42-3 


444 


46-0 


43-2 



Table X. 

Abstract of Observations made during Summer, East Coast of Graham's Land, Weddell and Biscoe Seas. 
Mean Barometric Pressure at 32° Sea-Level and Standard Gravity at Latitude 45°. 



Position. 


Expedition. 


Lat. 

S. 


Long. 


Month. 


Days of 
Observa- 
tions. 


Baro- 
metric 
Pressure. 


Mean Air 
Tempera- 
ture. 


Mean Sea 
Surface 
Tempera- 
ture. 


Mean 
Cloud. 


Mean 
Force 
Wind. 






. , 


o , 






Inches 


o 


= 


0-10 


0-12 


E. coast of Graham's Land 


Ross, 1842 


62 47 


53 12 W. 


Dec. 


8 


29-113 


33-1 


32-7 




4 







,, 1843 


64 22 


56 13 W. 


Jan. 


31 


29-116 


30-8 


30-8 




3 


2 




Bruce, 1892 ) 






(Dec. 


15 


29-402 


81 '1 




9-0 


2 


6 




1893 \ 


63 40 


55 00 W. 


-[ Jan. 


31 


29 318 


31-1 




9-3 


4 


1 




,. 1893 J 






I Feb. 


18 


29-205 


29-6 




9-3 


5 


3 




Donald, 1892 "j 






(■Dec. 


15 


29-403 


32-4 


32-0 










1893 [ 


63 25 


56 00 W. 


A Jan. 


31 


29-321 


33-2 


32-0 










1893 j 






{ Feb. 


18 


29-242 


30-6 


313 








Weddell Sea . 


Ross, 1843 


63 49 


36 36 W. 


Feb. 


28 


29-303 


307 


30-7 




3 


6 






Bruce, 1903 . 


64 00 


29 45 W. 


Feb. 


28 


29-112 


27-8 


30-0 


91 


1 


8 






„ 1903 . 


64 01 


40 52 W. 


Mar. 


31 


29-035 


25-0 


29-4 


9-5 


2 


6 






,, 1904 . 


70 18 


17 47 W. 


Mar. 


31 


29-066 


25-0 


29-7 


8-8 


2 


8 






Ross, 1843 


69 06 


14 32 W. 


Mar. 


11 


29-122 


28-9 


30 2 




4 


9 


Biscoe Sea 




Moore, 1845 . 


64 09 


39 14 E. 


Feb. 


27 


29138 


32-3 


32-5 




4 











61 46 


85 14 E. 


Mar. 


10 


28-980 


33-5 


32-5 




3 


9 






"Challenger," 1874 


64 01 


86 21 E. 


Feb. 


17 


28-931 


30-7 


32-1 


9-0 


3 


9 


* Staten Island 


1892 


54 23 


63 47 W. 


Dec. 


15 


29-226 


46-0 




7-5 








1893 


54 23 


63 47 W. 


Jan. 


31 


29-432 


48-4 




6-8 






ii ii 


1893 


54 23 


63 47 W. 


Feb. 


18 


29-525 


45-9 




6-4 







* These observations are for the same periods as those included in the observations taken in 1892-93 by Bruce and Donald. 



Table XI. 

Aiistract of Meteorological Observations for February and March in the Weddell and Biscoe Seas, from 
Observations made on the "Scotia," "Challenger," "Pagoda," "Erebus," and "Terror." Barometric 
I'nE.ssruE at 32° Sea-Level and Standard Gravity. 















Latitude. 

S. 


Mean Position. 


Days. 
Observa- 
tions. 

45 
45 
18 
13 
13 
9 
28 
23 
51 


Mean 

Barometric 

Pressure. 


Mean Air 
Temperature. 


Mean 

Sea Surface 

Temperature. 




Lat. S. 


Long. 


K-hruary 
ii 

March 

ii 

• • 












o 

60 to 65 
60 „ 65 
65 ,, 70 
65 „ 70 
60 ,, 65 
60 ,, 65 
65 ,, 70 
70 „ 74 
65 ,, 74 


62 25 
62 42 

68 09 

66 20 
62 49 
61 57 

67 54 
72 23 

69 55 


37 23 W. 

63 45 E. 

22 43 W. 
49 52 E. 
30 56 W. 
84 12 E. 

23 09 W. 

20 03 W. 

21 45 W. 


Inches. 
29-258 
29-031 
29-223 
29-159 
29 099 
28-992 
29-141 
29-048 
29-099 


30-4 
326 
26-1 
30-2 
26-8 
33-4 
27-7 
23-6 
25-9 


30-9 
32-6 
29-2 
30-1 
30-0 
32-5 
30-0 
29-5 
29-8 



THE WEDDELL QUADRANT AND ADJACENT AREAS. 



129 



Table XII. 

Abstract of Observations made in the Ross Sea and Vicinity on Board the "Erebus" and 
Summer Barometric Pressure at 32° Sea-Level and Standard Gravity. 



Terror. 



Position S. 


Mean Lat. S. 


Mean Long. 


Days. 
Observations. 


Mean Barometric 
Pressure. 


Mean Air 
Temperature. 


Mean Sea Surface 
Temperature. 


60-00 to 62-30 
65-00 ,, 67-30 
67-30 ,, 75-00 
75-00 ,, 78-00 


60 30 
66 26 
71 16 
76 45 


143 40 W. 
156 46 W. 
177 09 E. 
177 37 E. 


8 
42 
55 
34 


Inches. 
29-148 
29-289 
28-958 
29-039 


34-8 
308 
27-8 
25 9 


35-9 

28-8 
30-0 
29-7 



Table XIII. 

Showing the Percentage Frequency and Mean Direction of Wind in the Weddell Quadrant, Biscoe Sea, 

and Ross Sea during Summer and Early Autumn. 



Locality. 


Lat. 

S. 


Long. 


Observers. 


Months. 


Days. 
Obser. 


N. 


N.E. 


E. 


S.E. 


S. 


S.W. 


W. 


N.W. 


Mean. 










vations. 




















E. Coast of Graham's Land . 


63 48 


55 13 W. 


Bruce and Ross 


Dec.-Feb. 


103 


12 


14 


10 


12 


18 


12 


9 


13 


S. 22 E. 




63 40 


55 00 W. 


Bruce 




64 


13 


15 


11 


14 


20 


6 


7 


14 


S. 50 E. 






63 25 


56 0U W. 


Donald 




64 


13 


13 


15 


12 


24 


2 


13 


8 


S. 70 E. 






64 22 


56 13 W. 


Ross 


Jan. 


31 


11 


15 


10 


8 


13 


28 


12 


3 


S. 25 W. 


Weddell Sea 




63 49 


36 36 W. 


Ross 


Feb. 


28 


9 


10 


18 


28 


18 


11 


2 


4 


S. 44 E. 










70 24 


21 59 W. 


Ross and Bruce, S. of 67° 30' 


March 


43 


16 


20 


19 


9 


5 


7 


8 


16 


N. 18 E. 










68 01 


9 43 W. 


Biscoe and Bruce 


Jan. and Feb. 


31 


6 


6 


14 


45 


13 


13 


2 


1 


S. 38 E. 


Biscoe Sea 








64 09 


39 14 E. 


Moore 


Feb. 


27 


13 


24 


17 


9 


2L 


7 


3 


6 


N. 60 E. 










63 36 


50 10 E. 




Feb. and Mar. 


37 


10 


22 


17 


IS 


20 


9 


4 


5 


N. 81 E. 










6130 


59 48 E. 


„ 60°-62°30' S. 




15 


3 


7 


11 


13 


35 


20 


7 


4 


S. 42 E. 


, 








64 52 


44 18 E. 


„ 62i°-67° S. 




22 


17 


32 


21 


13 


9 


1 


1 


6 


N. 58 E. 










66 08 


41 16 E. 


Biscoe 




52 


2 


1 


11 


18 


44 


12 


8 


4 


S. 64 E. 


,, 








60 06 


68 11 E. 




Mar. and Apr. 


23 


16 


52 


12 


4 


9 


2 


5 





N. 50 E. 


, 








64 01 


86 21 E. 


"Challenger" 


Feb. 


17 


10 


4 


15 


17 


14 


15 


12 


13 


s. ;w, 


South Atlantic 






59 34 


27 29 W. 


Various 56°-64° S. 


Dec.-Feb. 


45 


10 


4 


3 


3 


18 


30 


15 


17 


S. 72 W. 








58 30 


43 22 W. 


„ 56°-64° S. 




30 


3 








1 


6 


24 


42 


24 


S. 88 W. 








53 19 


58 53 W. 


„ 52°-56° S. 




19 


7 


14 


2 


1 


2 


27 


27 


20 


S. 79 W. 


,, 






46 52 


7 59 W. 


"Scotia" 


Apr. 


30 


15 


7 


4 


4 


4 


14 


21 


31 


N. 58 W. 


*Staten Island 






54 23 


63 47 W. 




Dec.-Feb. 


64 


11 








5 


26 


16 


20 


22 


S. 88 W. 


Ross Sea Region 






60 30 


143 40 W. 


Ross 60°-62£° 




8 


4 








1 


27 


32 


24 


12 


S. 78 W. 


ii ii i> 






66 26 


156 46 W. 


,, 65"-67i° 




42 


11 


17 


13 


15 


16 


10 


8 


10 


S. 75 E. 


ii n ii 






71 16 


177 09 E. 


„ 67$°-75° 




55 


7 


6 


8 


15 


19 


18 


17 


10 


S. 9W. 


,. i. 






76 45 


177 37 E. 


,, 75°-78° 




34 


8 


7 


10 


22 


22 


18 


9 


4 


S. 10 E. 



* These observations are for the same 64 days as those included in the observations of Drs Bruce and Donald given above. 



Table XIV. 

Mean Force of the Wind, Scale 0-12, in the Weddell Quadrant, Biscoe Sea, and Ross Sea during 

Summer and Early Autumn. 

















Mean Force 








Position. 


Lat. 
S. 


Long. 


Observers. 


Months. 


No. of 
Obser- 




































vations. 


N. 


N.E. E. 

1 


S.E. 


S. 


S.W. 


W. 


N.W. 


Mean. 


E. coast of Graham's Land . 


63 48 


55 13 W. 


Ross and Bruce 


Dec.-Feb. 


641 


2-4 


3-3 3-0 


3-1 


53 


3-9 


29 


3-3 


3-5 


ii ii 


63 40 


55 00 W. 


Bruce 




433 


3-9 


3-3 34 


34 


60 


5-1 


?.-4 


32 


4-0 


ii ii 


62 47 


5312 W. 


Ross 


Dec. 


41 


1-5 


2-0 .. 




4-3 


4-7 


5-5 


3-6 


4-0 


ii ii 


64 22 


56 13 W. 


Ross 


Jan. 


180 


46 


33 i 20 


2-0 


30 


3-3 


3-1 


4-1 


3-2 


Weddell Sea ... 


63 49 


36 36 W. 


Ross 


Feb. 


164 


3-9 


2-7 1 3-4 


4-2 


39 


39 


3-5 


39 


36 


ii ii 






66 17 


28 03 W. 


"Scotia" 


Feb. and Mar. 


2059 


2-4 


2-5 33 


?,•?, 


1-9 


2-1 


2-9 


24 


2-5 


ii ii 






69 06 


14 32 W. 


Ross 


Mar. 


63 


3"2 


62 63 


4-7 


24 


1-5 




2-5 


49 


ii ii 






70 24 


21 59 W. 


Ross and " Scotia," S. of 67° 30' S. 




1026 


2-5 


35 4-5 


3-8 


T7 


2-1 


2-7 


2 4 


5 


Biscoe Sea . 






64 07 


39 14 E. 


Moore 


Feb. 


151 


3-5 


37 3-4 


4-7 


5'6 


4-2 


37 


4-9 


40 


ii ii 






63 36 


60 10 E. 




Feb. and Mar. 


212 


3-4 


35 I 3-7 


4-3 


5-6 


4 1 


4-0 


39 


39 


ii ii 






61 30 


59 48 E. 


„ 60°-62J° S. 




88 


2-8 


3-3 26 


3 5 


5 


4-1 


41 


3-5 


3 9 


ii ii 






64 52 


44 13 E. 


,, 62£°-67°S. 




124 


35 


39 1 3-6 


4 9 


63 


4-5 


3-7 


38 


4 


i. ii 






64 01 


86 21 E. 


" Challenger " 


Feb. 


392 


4-5 


28 ! 3-6 


4-91 


3-7 


2-7 


3-4 


4-4 


39 


Ross Sea Region 






60 30 


143 40 W. 


Ross 60°-62£° 




48 


5-2 


1 •■ 


5 


4-0 


5-9 


5fi 


5-4 


5-2 


ii ii ii 






66 26 


156 46 W. 


,, 65°-67°30' 




245 


3-7 


30 1 2-7 


2-4 


3-3 


3-4 


3-7 


35 


30 


ii ii » 






7116 


177 09 E. 


,, 67°30'-75° 




319 


33 


3-1 4-5 


5-2 


4-7 


42 


47 


4-6 


4-5 


"""•■■ 


76 45 


177 37 E. 


„ 75°-78* 




198 


30 


31 3-4 5-3 


4-7 


4-5 


3-5 


3-5 


4-3 



130 



MB ROBERT C. MOBSMAN ON THE METEOROLOGY OF 



Table XV. 
Mean Force of the Winds, " Belgica," Houks 4-8 Noon, 4-8 Midnight. (Scale 0-12. 



Month. 



January 

February 

March 

April 

May 

June 

July 

August 

September 

October 

November 

December 

Year . 

Spring 
Summer 
Autumn 
Winter 



Mean Position. 



Lat. S. Long. W. 



70 13 

70 40 

71 03 

70 67 

71 04 
71 09 
70 44 
70 28 
70 07 
70 04 
70 05 
70 04 
70 35 

70 06 

70 19 

71 01 
70 47 



86 56 

92 40 

93 10 
90 12 
88 58 
86 40 

86 40 

87 20 
82 44 

81 40 

82 14 
84 18 
87 22 

82 13 
87 58 
90 47 
86 33 



2-8 
2-0 
2-7 
3-0 
2-9 
1-8 
1-6 
2-9 
31 
2-8 
2-2 

2 V 7 

2-8 
22 
29 
23 



N.E. 



31 
2-9 
4-7 
3-3 
33 
2-5 
2-7 
2-0 
31 
2-4 
30 
28 
3-1 

29 
29 
38 



E. 


S.E. 


S. 


25 


2-1 


2-6 


3-5 


36 


2-7 


62 


2-4 


IS 


. 4-8 


52 


3-0 


3-2 


36 


2'4 


2-5 


2-6 


1-9 


2-0 


2-9 


1-7 


2-4 


2-4 


2'5 


2-6 


2-1 


1-6 


35 


39 


3-2 


34 


30 


1-6 


28 


2-6 


1-7 


32 


31 


2-1 


3-2 


2-7 


22 


2-9 


2-8 


2-1 


4-8 


4-2 


2-3 


2-2 


2-7 


1-8 



s.w. 



w. 



1-6 
2-4 
2-2 
2-0 
2-3 
2-6 
1-7 
2'3 
1-2 
3-2 
1-4 
2-7 
2-2 

2-2 
2-4 
2-2 

2-2 



1-3 
1-9 
2-8 
3-2 
2-6 
28 
1-9 
2-5 
2-9 
3-3 
2-3 
2-6 
2-7 

2-9 

2-2 
2-8 
26 



* Includes March 1898 and March 1-14 1899. 



Table XVI. 



N.W. 


Mean. 




2-7 


1-6 


2-9 


2-8 


3'2 


3-2 


38 


2-5 


2-8 


20 


2-5 


2-5 


2-0 


3 


26 


3 


2-5 


2'3 


30 


2 '3 


2-6 


10 


2-6 


26 


2-6 


2-6 


2-7 


1-3 


2-7 


2-8 


33 


2-7 


2-4 



Gales, Force 8 or upwards, recorded at Dungeness (Lat. 52° 24' S., Long. 68° 25' W.) and 
Evangelists Island (Lat. 52° 24' S., Long. 75° 06' W.), 5 Years 1901-05. 



All Gales recorded Percentage Frequency, Dungeness. 


All Gales recorded Percentage Frequency, Evangelists Island. 




N. 

5 
1 
6 
3 
4 


N.E. 

1 
1 

3 
I 


E. 



2 
2 
1 


S.E. 




2 

1 


S. 

2 
1 
2 
9 
4 


S.W. 

69 
92 
77 
53 
73 


w. 

18 

4 

7 

20 

12 


N.W. 

5 
1 
3 

8 
4 




N. 

4 

11 

11 

5 

7 


N.E. 








E. 




7 
2 


S.E. 


1 
1 
1 
1 


S. 

1 

2 
12 
12 

7 


S.W. 

10 
13 
12 
18 
13 


w. 

20 
20 
18 
17 
19 


N.W. 


Spring . 
Summer 
Autumn 
Winter . 
Year 




Spring . 
Summer 
Autumn 
Winter . 
Year 


65 
53 
46 
40 
51 




Gales per 1000 days of each Wind, Dungeness. 


Gales per 1000 days of each Wind, Evangelists Island. 




N. 


N.E. 


E. 


S.E. 


S. 


S.W. 


W. 


N.W. 


Mean 




N. N.E. 


E. 


S.E. 

37 
125 
35 
29 
37 


S. 

19 

34 

152 

121 

96 


S.W. 


w. 


N.W. 


Mean. 

157 
115 
148 
141 
140 


Spring . 

Summer 
Autumn 
Winter. 
Year 


87 
28 
85 
41 
63 


15 
32 

68 
31 






95 

75 
50 





103 

62 
46 


79 
55 
56 

i r.> 

88 


192 
201 
151 
136 
172 


113 
36 
35 
99 
74 


44 
16 
23 
75 
38 


108 

114 

85 

92 

99 


Spring 
Summer 
Autumn 
Winter 
Year . 


73 
152 
161 

76 
118 















146 

86 


113 

87 
119 
153 
118 


131 
92 
131 
152 
124 


281 
177 
273 
245 
242 




Monthly Gale Frequency per Cent. 




Jan. 

9-5 
147 


Feb. 


Mar. 


April. 


May. 


June. 

7-8 
15-3 


July. 


Aug. 


Sept. 


Oct. 


Nov. 


Dec. 


Year. 

9-9 
14-0 


Dungeness . 
Evangelists Isl 


and 


• 


• 


1 

] 


6 

11-6 


11-9 
157 


6-0 
16-2 


7-6 
12-3 


9-7 
11-4 


10-2 
15'5 


124 
14-7 


11 
17-3 


87 
14 7 


9-1 

8-7 



THE WEDDELL QUADRANT AND ADJACENT AREAS. 



131 



Table XVII. 
Mean Barometric Pressure at 32° Sea-Level and Standard Gravity; Mean Temperature; Cloud Amount; 
and Wind Force, with Various Winds, at Laurie Island, South Orkneys, and Cape Pembroke, Falkland 
Islands, for the Seven Months April to October 1903. 





South Orkneys. 






Cape Pembroke, Falkland Islands. 


Wind. 


Pressure. 


Temperature. 


Cloud. 


Wind Force. 


Wind. 


Pressure. 


Temperature. 


Cloud. 


Wind Force. 




Inches. 





0-10 


0-12 




Inches. 


o 


0-10 


0-12 


N. 


29-351 


24-8 


9-1 


25 


N. 


29-712 


419 


7-5 


4-0 


N.E. 


■208 


22-4 


9 5 


1-6 


N.E. 


•808 


41-4 


8-5 


4-9 


E. 


•250 


8-1 


6 6 


1-2 


E. 


•804 


38'6 


4'8 


2-9 


S.E. 


•234 


5-6 


7-2 


1-3 


S.E. 


■725 


34-9 


7-0 


3-6 


S. 


•215 


5-2 


6-5 


1-3 


S. 


•780 


37 2 


77 


4 3 


s.w. 


•318 


7-4 


6-3 


1-4 


S.W. 


■634 


37-6 


7-5 


5 1 


w. 


•215 


23-9 


7-0 


1-9 


w. 


•649 


37 9 


5-8 


4-2 


N.W. 


•320 


26-9 


8-6 


2-8 


N.W. 


•714 


400 


5-9 


3-9 


Calm 


■386 


11-0 


57 




Calm 










Mean 


29-297 


17-6 


7-6 


2-0 


Mean 


29-715 


39-8 


6-7 


4-1 



Table XVIII. 

Mean Barometric Pressure at 32° Sea-Level and Standard Gravity, Mean Temperature, Cloud, Wind Force, 
with Various Winds, from Observations made at Joinville Land and "Erebus and Terror Gulf" by 
Dr W. S. Bruce on board the "Bal^ina," December 16, 1892, to February 18, 1893. Mean Latitude 
63° 40' S., Mean Longitude 55° 00' W. 



Wind Direction and Number of Observations. 


Pressure. 


Temperature. 


Cloud. 


Wind Force. 




Observations. 


Inches. 





0-10. 


0-12. 


N. 


55 


29-434 


31-9 


8-6 


3-9 


N.E. 


62 


•375 


31-4 


9 3 


3-3 


E. 


46 


•501 


31'2 


8-7 


3-4 


S.E. 


61 


■145 


29-8 


9-1 


3-4 


S. 


85 


•171 


27-8 


9-2 


60 


S.W. 


23 


•163 


31-4 


9 2 


5-1 


w. 


30 


•326 


33 2 


7-0 


2-4 


N.W. 


58 


•306 


32-4 


9-0 


3-2 


Calm 


55 


•486 


31-9 


8-0 




Mean 


48 


29-308 


30-8 


9-2 


4-0 



Table XIX. 

Showing the Departure from the Average of Mean Values of the Climatic Elements in the Weddell 
Sea with Different Winds South of 60° S. during the Cruises of the "Scotia" in 1903-04. 

Note. — The heavy type indicates above the average, and the italic type below the average. 



Mean Position. 


Wind 
Direction. 


Pressure. 


Temp. 


Cloud. 


Vapour 
Pressure. 


Humidity. 


Wind 
Force. 


No. of Observations. 






Inches. 


° 


0-10. 


Inches. 


Per cent. 


0-12. 




Per cent. 




IS. 


0119 


2 


0-4 


012 


1 


0-4 


212 


17 




N.E. 


•089 


07 


0-2 


001 


1 





242 


19 


1903. 


E. 


■134 


10 


07 


•001 


1 


0-7 


119 


9 


Lat. 64" 23' S. 

Long. 34° 37' W. 

53 days. 


S.E. 


■229 


1-6 


o-o 


■009 


1 


1 


117 


9 


Is. 


■057 


4-0 


0-5 


■020 


1 


0-2 


126 


10 


S.W. 


026 


5 


0-4 


■002 


3 


OS 


144 


11 


Feb. and March. 


w. 


■058 


1-0 


0-8 


■005 


3 


07 


109 


9 




N.W. 


■116 


10 


0-5 


015 


3 


03 


133 


10 




Calm 


•079 


1-9 


V3 


■014 


4 




70 


6 




\ Mean 


29-090* 


26-8 


9-4 


•123 


83 


2-4 


Total 1272 


100 




' N. 


0-172 


3 6 


0-8 


•019 


8 


0-3 


128 


14 




N.E. 


■080 


0-4 


1 


•010 


6 





102 


11 


1904. 


E. 


■114 


2-3 


0-3 


•010 


6 


0-8 


100 


11 


Lat. 70° 03' S. 


S.E. 


121 


1-0 


0-2 


■002 


2 


1-1 


52 


6 


Long. 20° 43' W. 


S. 


•229 


3-9 


1-2 


■027 


10 


1-4 


66 


7 


38 days. 


S.W. 


■112 


0-4 


0-8 


■018 


15 


0-2 


131 


14 


Feb. and March. 


W. 


•154 


1-9 


0-5 


005 





o-o 


136 


15 




N.W. 


•040 


0-6 


5 


008 


5 


01 


142 


16 




Calm 


•299 


4-0 


V8 


•035 


4 




55 


6 




Mean 


29-018* 


25 7 


87 


•118 


83 


2-6 


Total 912 


100 



At 32° Sea-Level and Standard Gravity at Lat. 45" 



132 



M R ROBERT C. MOSSMAN ON THE METEOROLOGY OF 



Table XX. 
Tiikiim.u. and Barometric Windkose During the Drift of the "Belgica" foe the Seasons and the Year. 



Meau Temperature of the Winds. 


N. 


N.B. 


E. 


S.E. 


s. 


S.W. 


w. 


N.W. 


Calm. 


Mean. 




„ 


o 


o 


o 




















Spring .... 


26'2 


20-3 


13-6 


1-4 


- 3-3 


5-2 


10-4 


21-7 


5-5 


12-0 


Summer .... 


81 "1 


31-6 


31 -0 


28-9 


24-9 


25-9 


26-6 


29-2 


26-6 


29-3 


Autumn .... 


25-0 


24-6 


18-8 


7-5 


8-1 


8-4 


12-2 


19-1 


7-9 


15-6 


Winter .... 


11-5 


12 6 


11-8 


8-1 


- 6-5 


- 9-9 


2'2 


11-5 


-12-1 


1-7 


Year 


23 9 


24-6 


22'2 


15-7 


5-5 


5-2 


9-0 


17-6 


4-3 


14-6 






Departure from Annual Mean.* 














N. 


N.E. 


E. 


S.E. 


S. 


S.W. 


W. 


N.W. 


Calm. 


Mean. 




„ 


o 


o 


o 



















Spring .... 


2-3 


4'S 


8-6 


14-3 


8-8 


o-o 


1-4 


41 


1-2 


2-6 


Summer .... 


7 2 


7-0 


8-8 


13-2 


19-4 


20-7 


17 6 


11-6 


22 3 


147 


Autumn .... 


11 


o-o 


3-4 


8-2 


2-6 


3 2 


3 2 


1-5 


3-6 


1-0 


Winter .... 


12-4 


12-0 


W4 


7-0 


12-0 


15-1 


6-8 


6-1 


16'4 


12-9 






Mean Barometric Pressure of the Winds. 












N. 


N.E. 


E. 


S.E. 


S. 


S.W. 


W. 


N.W. 


Calm. 


Mean. 


Spring .... 


9-225 


9-300 


9-221 


9158 


9-178 


9-398 


9-438 


9 351 


9-382 


29-316 


Summer .... 


9012 


9-394 


9-288 


9-166 


9-288 


9-315 


9-296 


9-296 


9-398 


29-277 


Autumn .... 


9-343 


9-091 


9-012 


9-036 


9-347 


9-162 


9-288 


9-382 


9-272 


29-194 


Winter .... 


9-461 


9-347 


9-233 


9-180 


9-300 


9-339 


9-611 


9-489 


9-528 


29-442 


Year 


9-288 


9-288 


9-205 


9-142 


9-221 


9-312 


9-461 


9-406 


9-410 


29 307 


Departure from Annual Mean.* 


N. 


N.E. 


E. 


S.E. 


S. 


S.W. 


W. 


N.W. 


Calm. 


Mean. 




Iris. 


Ins. 


Ins. 


Ins. 


Ins. 


Ins. 


Ins. 


Ins. 


Ins. 


Ins. 


Spring .... 


■063 


012 


•016 


•016 


■043 


•086 


•023 


■055 


■028 


009 


Summer .... 


■276 


•106 


•083 


024 


067 


•003 


•165 


■110 


■012 


■030 


Autumn .... 


055 


■197 


■193 


■106 


126 


■150 


•173 


■024 


■138 


■113 


Winter .... 


•173 


059 


028 


038 


•079 


•027 


•150 


•083 


118 


135 



The heavy type indicates an excess, and the italic type a defect. 



Table XXI. 

Thermal Windrose, Staten Island and Rawson. Departure from the General Mean of the 

Months specified. 

Staten Island. Latitude 54* 23' S., Longitude 63° 47' W. 





N. 


N.E. 


E. 


S.E. 


s. 


S.W. 


w. 


N.W. 


January .... 


+ 1-4 


-1-8 


- 0-4 


-1-4 


-0-9 


-0-2 


+ 1-3 


+ 2 2 


April 


+ T4 


+ 1-8 


-0-4 


-2-0 


-3-2 


-1-4 


+ 2-0 


+ 1-8 


Julv .... 


+ 2-5 


+ T6 


+ 1-1 


-2-2 


-4-0 


-1-1 


+ 0-2 


+ 2-0 


October .... 


+ 2-0 


+ 1-8 


-1-1 


-3-8 


-3-2 


+ 0-2 


+ 1-4 


+ 2-7 


Year 


+ 1-8 


+ 0-9 


-0-2 


-2-3 


-2-9 


-0-5 


+ 1-3 


+ 2-2 




Rawson 


Latitude 43° 17' S., 


Longitude 65° 05' W. 








J.muary .... 


+ 2-9 


-3-4 


-1-4 


-1-8 


-3-1 


+ 07 


+ 2-2 


+ 4-1 


April .... 


+ 1-8 


+ 0-2 


+ 0-5 


-1-3 


-36 


+ 0-2 


-0-7 


+ 2-7 


July 


+ 1-8 


+ 2-2 


o-o 


-1-8 


-25 


-4'3 


-0-2 


+ 4-7 


October .... 


+ 3-1 


+ 1-3 


-2*5 


-4-3 


-5-4 


+ 1-1 


+ 3-2 


+ 3-6 


Year 


+ 2-3 


o-o 


-0-9 


-2-3 


-3-6 


-0-5 


+ 11 


+ 3-8 



THE WEDDELL QUADRANT AND ADJACENT AREAS. 



133 



Additional Note, received 6th September 1908. 

Note on the Mean Monthly and Annual Barometric Pressure at Sandy 
Point, the Falkland Islands, and in the Cape Horn Regions. 

The maps of mean monthly pressure given in this paper may he said to approxi- 
mately represent the conditions prevailing during the five years 1903-1907. For this 
reason the trend of the isobars over the ocean in the Cape Horn region as well as their 
relative height will be found in some months to differ from those given in treatises 
dealing with the results of maritime observations obtained from ships' logs extending 
over long terms of years. With a view of giving good normals for land stations in this 
region, I have made every effort to obtain values for Sandy Point, the Falkland Islands, 
and for the Cape Horn area. For Cape Pembroke (Falkland Islands) the values of mean 
monthly and annual pressure which follow are based on 16|- years' observations. For 
the vicinity of Cape Horn, means deduced from observations taken at Staten Island, 
Ushuaia, Harbertown, and New Year's Island are given. The periods embraced by the 
records comprise 19 years for the months of February, March, and May, 20 years for 
January, April, June, August, and December, 21 years for July and September, and 22 
years for October and November. The monthly averages were obtained by taking the 
mean of all or any of the above stations, observing simultaneously. Thus for the year 
1905 the monthly means are the average of the mean pressures recorded at New Year's 
Island, Ushuaia, and Harbertown. By this method a much closer approximation to the 
true pressure was obtained than if the monthly means had depended on only one 
station. The mean position of the stations is 54^ degrees south latitude, and 65 degrees 
west longitude. Their range in latitude does not exceed a few miles, but in longitude 
the stations differ some 4|- degrees or 160 miles. This variation in longitude is, how- 
ever, unimportant, as the isobars in this locality normally approximate to parallels of 
latitude. For Sandy Point the monthly means are based on 18 years' observations, so 
that the values for all three districts are very nearly of the same length, although they 
are not quite synchronous, being for different terms of years. In spite of this deficiency, 
however, owing to the paucity of land observations in this part of the globe, they 
afford the closest approximation that it is at present possible to obtain for the distribu- 
tion of pressure at the south of the continent of South America, and over the ocean to 
the east up to the longitudes of the Falklands in about 52° S. latitude. The follow- 
ing table shows the mean monthly and annual pressure at the three stations. 

Mean Monthly Barometric Pressure at 32" Sea-Level and Standard Gravity. 



Station. 


Period. 


Years specified. 


Jan. 


Feb. 


Mar. 


Apr. 


May. 


June. 


July. 


Aug. 


Sept. 


Oct. 


Nov. 

29'520 
•434 

•559 


Dec. Year. 


Sandy Point . 

Cape Horn ) 
Region . j 

C. Pembroke or 
Port Stanley 
(Falklands) . 


18 years 
19-22,, 


1 1871-72, 1889- \ 
\ 1902, 1905-06 / 
( 1876-79, 1881-83, ) 
1 1886-96, 1899, } 
( 1902-07 J 

/ 1859-68, 1875-77, \ 
\ 1882-83. 1903-05 / 


29-485 
■&31 

481 


29473 
-445 

■531 


•29-525 
■432 

■AM 


29-552 
•502 

•539 


29-596 
'495 

•538 


29-651 
•551 

■606 


29-571 
•527 

■576 


29-644 
•570 

•549 


29650 
■625 

-644 


29-599 
•538 

•623 


■438 
•492 


29-559 
■499 

•551 



134 



MR ROBERT C. MOSSMAN ON THE METEOROLOGY OF 



As regards the Cape Horn region and Sandy Point, the above means differ but little 
from those o-iven in Table II. At the Falkland Islands, however, the differences are 
considerable, amounting on the mean of the year to an excess of 0*067 inch during the 
short period, while in July, August, November, and December the excess amounts to 
more than a tenth of an inch. At all three stations it will be observed the highest 
mean monthly pressure occurs in September. The month of minimum pressure, on the 
other hand, is December at Sandy Point, January and March for the Cape Horn area, 
and March at the Falklands. The range in the mean monthly pressures is 0*204 inch at 
Sandy Point, 0*194 inch in the vicinity of Cape Horn, and 0*170 inch at the Falkland 
Islands, being greatest in the region most removed from the direct influence of the 
ocean, and least at the strictly insular station on the Falklands. It is of interest to 
give for the Cape Horn area the difference of mean monthly pressure for the five years 
1903-1907, as compared with the above long average, the plus sign indicating that 
pressure during these five years was in excess, and the minus sign that it was in 
defect, of the long average. 

July. Aug. Sept. Oct. Nov. Dec. Year. 

in. in. in. in. in. in. in. 

+ •084 --012 --114 +-007 -'015 +-014 -"001 

Thus during the past five years pressure in the Cape Horn region has on the mean 
of the year been virtually the average, but with considerable variations in the months, 
ranging from an excess of + "084 inch in July to a defect of — 0*114 inch in September. 

The following table showing the highest and lowest mean monthly pressures in the 
vicinity of Cape Horn may be of interest, owing to the comparatively (for the region) 
long period covered by the combined records. 



Jan. 


Feb. 


March. 


April. 


May. 


June. 


in. 


in. 


in. 


in. 


in. 


in. 


+ •058 


+ ■066 


-•080 


+ -040 


+ ■032 


-•098 



Capb Horn Region Latitude 54° 30' S., Longitude 65° W. Highest and Lowest Mban 
Monthly Pressure at 32° Sea-Level and Standard Gravity. 









Highest. 


Lowest. 


Range. 




Inches. 




Inches. 




Inches. 


January 
February 
March . 
April 
May 
June . 
July . 
August 
September 
October 
November 
December 
'Year . 






29-752 
•610 
•840 
•811 
•737 
•816 
•723 
•897 
•890 
•876 
•759 
•722 

29-646 


1878 
1896 
1878 
1878 
1904 
1888 
1887 
1907 
1886 
1876 
1876 
1904 
1878 


29-236 
170 
•185 
•310 
•284 
•285 
•287 
•342 
•353 
•256 
T22 
•210 

29412 


1890 
1891 
1895 
1892 
1907 
1905 
1889 
1905 
1889 
1892 
1896 
1889 
1889 


0-516 
0-440 
0-655 
0-501 
0-453 
0-531 
0-436 
0-555 
0537 
620 
0-637 
0-512 
0-234 



* The years with complete observations are 1876, 1878, 1888-1893, 1896, 1903-1907. 



THE WEDDELL QUADRANT AND ADJACENT AREAS. 135 

Hence the highest mean monthly pressure was 29 - 897 inches in August 1907 and 
the lowest 29 - L22 inches in November 1896, showing a range of - 775 inches. Looking 
at the last column, it will be seen that the greatest monthly range or difference between 
the highest and lowest mean monthly pressures is 0'655 inches in March, and the least 
0'436 inches in July. 



REFERENCES. 

Antarctic Manual, London, 1901. 

Arctowski, Henryk, Expedition Antarctique Beige: Resultats du Voyage du S.Y. " Belgica" en 1897- 

98-99 : " M^teorologie : Rapport sur les Observations M^teorologiques Horaires," Anvers, 1904. 
Wilson-Barker, D., Commander R.N.R., "The Present Position of Ocean Meteorology," Quar. Jour. 

Royal Meteorological Society, vol. xxx. pp. 105-132. 
Bodman, Gosta, " Meteorologisch e Ergebnisse der Schwedischen sudpolar Expedition," Petermanns Geogr. 

Mitteilungen, 1904, Heft v. 

" Om Klimatet i Antarktis mel. sarskild hainsyn till Graham Land, U R Ymer," Tidskrift 

Utgifoenaf Svenska Gd/lskapet for Antropologi och Geografi, Arg. 1904, H. 3. 

Bruce, W. S., " Bathymetrical Survey of the South Atlantic Ocean and Weddell Sea," Scottish Geographical 

Magazine, August 1905. 
Buchan, A., Report on Atmospheric Circulation. 
Chili : Annuario del Servicio Meteorologico de la Direccion del Territorio Maritimo, 1899-1905, Valparaiso, 

1902-1906. 
Contributions to our Knowledge of the Meteorology of the Antarctic Regions, published by the authority of the 

Meteorological Committee, London, 1873. 
" Challenger" Reports : "Narrative," vol. ii. 
Davis, Walter G., Climate of the Argentine Republic, Buenos Aires, 1902. 

Anales de la Oficina Meteorologica Argentina,, Tomo xi. 

"Observations made at Laurie Island, South Orkneys, in the Year 1904," Annals of the Argentine 

Meteorological Office, vol. xvi., in Spanish and English. 
Dobrowolski, A., Expedition Antarctique Beige ; Resultats du Voyage du S.Y. " Belgica" en 1897-98-99 : 

" M^teorologie : Observations des Nuages," Anvers, 1903. 
Gould, Benjamin A., Anales de la Oficina Meteorologica Argentina, Tomo iii. 
Hepworth, M. W. Campbell, C.B., Commander R.N.H., The Relation between Pressure Temperature and 

Air Circulation over the South Atlantic Ocean, published by authority of the Meteorological Com- 
mittee, Official No. 177, London, 1905. 
Lephay, J., Mission Scientifique du Cap Horn, 1882-1883, Tome ii., " Meteorologie," Paris, 1885. 
Marabini, P., Obseruatorio Meteorologico del Colegio Salesiano " S. Jose" en Punta Arenas de Magallanes 

(Chile): Resumende las observaciones de quince anos (1888-1902), Santiago de Chile, 1904. 
Meinardus, Wilh., Uber die Wind Verhdltnisse au der Winter station des " Gauss," Berlin, 1905. 
Mobsman, R. C, Results of the Scottish National Antarctic Expedition, vol. ii. of Report on the Scientific 

Results of the Voyage of the S.Y. " Scotia," Edinburgh, 1907. 
" Preliminary Reports on the Meteorological Results of the Scottish National Antarctic Expedition," 

Scottish Geographical Magazine, vol. xx. pp. 113-120; xxi. pp. 417-429; xxii. pp. 252-272. 
" The Recent Voyage of the 'Uruguay,'" Scottish Geographical Magazine, June 1905. 

"The South Orkneys in 1907," Scottish Geographical Magazine, July 1908. 

Rey, Lieut. J., "Note sur les Observations Meteorologiques " (Appendix to Le Francois au Pole Sud, by 

J. B. Charcot), Paris, n.d. 
South Georgia : Die Internationale Polarforsclmng , 1882-1883 : Die Beobachtungs-Ergebnisse der Deutschen 

Stationen, Band ii. : " Sud-Georgien," Berlin, 1886. 
Webster, W. H. B., Narrative of a Voyage to the Southern Atlantic Ocean in the Years 1828-29-30, 2 vols., 

London, 1834. 
Weddell, James, A Voyage toivards the South Pole, London, 1825. 

TRANS. ROY. SOO. EDIN., VOL. XLVII. PART I. (NO. 5). 19 



136 METEOROLOGY OF THE WEDDELL QUADRANT AND ADJACENT AREAS. 



NOTE. 

The maps on Plates 1. to IV. show average barometric pressures, temperatures, and amounts of cloud 
for the region dealt with in tliis paper. They are based on the values given in the foregoing tables, adjusted 
in some cases so as to indicate what is believed to be more nearly the true average (see page 104). The 
isobars, etc., on these maps were drawn by Mr E. K. Carmichael from those on sketch-maps supplied by 
the author. 

The general map, Plate V., is based on that accompanying a paper on the Bathymetrical Survey of the 
South Atlantic Ocean and Weddell Sea, by Dr W. S. Bruce, LL.D. (Scot. Geographical Mag., vol. xxi. 
p. 402, August 1905). The map has been revised by Dr Bruce for this paper. 



Trans. Roy. Soc. Edin. Vol. XLVII. 

Mr Robert C. Mossman on " The Meteorology of the Weddell Quadrant and Adjacent Areas." — Plate I. 







ES 


s s 

5 s 




1 \ 




a 

CD 
> 
O 

55 


/ ? 

a 

1 


8. 

I! 


i 


^ 






ml 




\V 1 N ' 


o 










v-.^jr--; 


« 




■fe^j 


w 


/ 


1 "tS 










/ 


i n 
i 


\ 








« 


I 


'. c 










I 


I o 


S 






o 









?? s s g 


i ' i 


J 

y Fig 2 


5 

1 

A 
* 


I 




— ^ 




^ 


/ 


1 ■ / i o 


G 


o 







P2 



Trans. Roy. Soc. Edin. Vol. XLVII. 

Mr Robert C. Mossman on "The Meteorology of the Weddell Quadrant and Adjacent Areas." — Plate II. 





1 


1 \ J 


oo 
E - 


C I 8 






M 


\ ■ j 


Q 


O 


i \ - 
j \ 

i .' 





1 






1 


00 

bb 


to 
a 
60 

3 


l 


l 


\ 1 
i 


| 


\ 




2 "j ? .--'''' 
'" 1 (f~~7~" 


*^ 




\ 


\[%C 


J-"\ 


% 


\ 


\% * ^ a 






| 


ItKTv" 










\ \\ 








1 




\ \\ 












'. o 












i o 


o 






s 




) 






1 


i 


60 " 


.V 


1 - X 




* 




1 /. 


o 


o 


! \ - 
) 













! 1 
! \ I 


Fig )( 
Octobt 


/■ 


1 


8 




X 






\ 














\A 


; 1 \ o 








f 


j*ijr~V' c 










\ 


I ' v n 












: \ 












1 ! c 












i \ - 


3 






o 




\ 







1 


y Fig- 


'1 


1 T 

Oft \ "\ o 




m 


: c 


C 


c 


! i S 
\ 









j \ 




E 




1 NJS 


60 <l 




I 


&* a. 


g 




0) 

I/} 


-w 




& ! 


Ik 




\ 




,-4 — ■ — ~" 


~N 


s 




\* 


i y-^ 


^\ 




\ \ 


V^&* <\ \ a 


I 












\ v$K 






\ \\ v < 








\\ N 








i |\\ 








i 1 = 










a 








"* 


<a 





Trans. Roy. Soc. Edin. Vol. XLVII. 

Mr Robert C. Mobsman on "The Meteorology of the Weddell Quadrant and Adjacent Areas." — Plate III. 






1 




00 « 

lib &o o» . 

/ / 1 


Y 


: T 




I 




) 

i / 

1 1 O 


c 


o 
u 


1 o 





1 

u 
fl) 


j 

! v = 




.o 

a 

£ * 8 

fl 1 

I \ A. 








1 ! O 


c 


o 









! \ 


£ ► 


= S .?" ? 

' / // 
/ f 


* J * 
/ /i 






r **4*; 


! e 


e 


c 


i i c 








| \ 


yf Fig. 6 


o 
a 
a .-». 


j 1 

i 

© { 

7 ; ® 

/ * / -^ 




^/ 






S 


; 







1 


v-/ Fig. 1( 


a 

2 ' 
o * 






'. 


1 1 

A^ 1 \ » o 








o 


c 


i o 





>-> 




i \ 
1 \ a 


<« 


_ 


8 ! ^ 


E § 

*~2 


/'/I 


( L^ 




s 


S I 


i a* s 








v%r 






^7 


\KS 






( 


I o 


i 


c 





1 






\ 


Fig 5 
May 






! \ e 




r SS 


I ^ 






'// 


/ / 






g 


( 


\*-% ; "'j o 








V 


Y - '^ ,> ' : 








**)) 


j\ 










\ © 










I o 


o 






o 


; 




Trans. Roy. Soc. Edin. Vol. XLVII. 

Mr Robert C. Mossman on "The Meteorology of the Weddell Quadrant and Adjacent Areas." — Plate IV. 






del. E. K. Carmichael. 



Fig. 1. — Isobars for the Year. Fig. 2. — Isotherms for the Year. Fig. 3. — Isonephs for the Year. 
Fig. 4.— Isobars and prevailing Winds for December 1904. Fig. 5. —Mean Surface-temperature of the Sea in Summer. 




Plate Y. 



138 MR R. C. MOSSMAN ON 

inches, with a maximum of 18"54 inches in June and a minimum of 2"85 inches in 
January. To the north of this wet zone the rainfall diminishes rapidly, while to the 
south precipitation continues high. No reports are available for the 600 miles of coast 
separating Ancud from Evangelists Island at the Pacific entrance to the Straits of 
Magellan, where the annual rainfall is 111 "44 inches, with a seasonal distribution the 
reverse of what obtains in the monsoonal region. 

At the island of Juan Fernandez, in lat. 33° 37' S., long. 78° 50' W., the distribution 
of the rainfall through the months of the year is in harmony with that prevailing on 
the Chilian littoral in the same latitude. The variation in the mean wind direction 
shows, however, little change ; being south in the month of January, when this region 
is well within the influence of the South Pacific high-pressure area, while in the 
month of June, when the high-pressure belt is north of its summer position, the winds 
at Juan Fernandez are south-west. These wind variations are also characteristic of 
the ocean region to the east of this island. 

To the north of latitude 30° S. the monsoonal effect on the littoral becomes masked 
or disappears ; but even at Caldera, in 27° S., the little rain that falls is precipitated 
in the rainy season of the more southern stations. At Iquique, situated in 20° S., 
where a shower of rain is of rare occurrence, the winters are much cloudier than the 
summers ; the seasonal amount of cloud on the mean of the three years 1901-03 being 
as follows : — 



Iquique. 


Cloud Amount 0-10. 


Years 1901-03. 




Hours of Observation 


8-2-9. 


Spring. 


Summer. Autumn. 


Winter. Year. 


6-4 


3-9 4-7 


72 5-5 



The indirect monsoonal effect would therefore appear to extend for some distance 
into the northern rainless reo-ion. 

O 

There can be little doubt that the influence of the mountain chain of the Andes 
in modifying the general circulation of the air is considerable, causing the coastal 
winds to conform to the shores of the littoral, and to blow parallel with the main axis 
of the Andean Cordilleras. The whole question clearly calls for a complete discussion 
of all the available Chilian data, and the co-ordination and correlation of these data 
with the observations from the Argentine Republic, Bolivia, Peru, and Brazil, as well 
as the joining up of the land values with those given in various memoirs dealing with 
the meteorology of the adjacent ocean regions. Among the factors bearing on the 
causation of the monsoonal effect may be mentioned the seasonal migration of the 
South Pacific anticyclone, the varying intensity of the high-pressure area in the north- 
west of Argentina, and the fluctuations in the southern atmospheric cyclonic centres 
of action in the Weddell and Bellingshausen Seas. The temperature gradient between the 
hind and the ocean at different seasons will also have important bearings on the solution 
of the whole question, and also variations in the temperature of the cold polar current 



THE MONSOONS OF THE CHILIAN LITTORAL. 139 

which flows along the west coast of South America, extending almost to Juan 
Fernandez. 

The monsoonal effect varies greatly from year to year ; some years being character- 
ised by weak monsoons, while in others the rain-bearing northerly air currents are 
greatly intensified. There appears to be a very definite interrelation between the 
position occupied by the Pacific and Atlantic " highs " and the Antarctic and sub- 
Antarctic circulation which may help to throw light on the varying weather conditions 
prevailing over the South American continent. To take merely a single instance, that 
afforded by a comparison of the conditions from 20° to 65° S. for July of 1903 and 
of 1904, based on returns from a large number of stations in South America and at 
the Falklands, as well as the sub-Antarctic stations of the South Orkneys, Snow Hill, 
and Wandel Island. In July 1903 the South Pacific " high," as shown by the winds and 
pressure at Juan Fernandez, was well south of its normal position, and rainfall was 
relatively small at all the Chilian coastal stations within the monsoonal area. In this 
region southerly winds were in excess, and northerly winds, which bring the rain, in 
defect of the average. Accompanying this excess of winds from higher latitudes, low 
temperatures prevailed over Argentina and Chili, except in the extreme south of the 
continent. Barometric pressure was below the normal at the South Orkneys and at 
Snow Hill on the east side of Graham's Land, the Weddell Sea "low" being not only 
deepened but lying farther west than usual. At Evangelists Island at the Pacific 
entrance to the Straits of Magellan the weather was mild and precipitation high, the 
same conditions extending to the Falklands and the South Orkneys, with westerly and 
north-westerly winds predominating. 

In July 1904, on the other hand, conditions were exactly the reverse of the month 
described. The South Pacific " high " was considerably to the north of the island of Juan 
Fernandez, where barometric pressure was 0*38 inch lower than in the July previous, 
while on the coast of Chili in the monsoon zone the fall of pressure averaged 0*20 inch 
as compared with the July of 1903. A permanent low-pressure area lay off the coast 
of Chili in lat. 42° S., as shown by the Daily Weather Maps of the Argentine Meteoro- 
logical Office. To the north-west of this region of low pressure the prevailing winds, as 
at Juan Fernandez, were from the west-south-west, while at coastal stations to the north 
of it the winds blew from the north and north-west. All over this region deluges of rain 
fell (see Table V.), the zone of heavy rainfall extending well into the centre of Argentina 
between latitudes 38° and 45° S., and over to the Atlantic coast. With this excess of 
equatorial winds, mean temperature was above the average. South of 45° S. the winds 
were southerly, blowing in on the northern " low." Temperature in the southern provinces 
of South America, at the Falklands, the South Orkneys, and Wandel Island was markedly 
below the normal, the winds in general circulating round the Weddell Sea " low," which 
was far to the east of its position in the July previous. Barometric pressure fell from 
west to east south of 55° S., and was as high as 29*51 # inches at Wandel Island on the 

* At 32°, sea-level and standard gravity. 



140 



MR R. C. MOSSMAN ON 



mean of the month, diminishing to 29 '33 inches at the South Orkneys, and still lower to 
the eastward. The month in its main features bore a very close resemblance to that of 
December 1904, which I have treated with some detail in the paper on the Weddell Sea 
meteorology. 

It will be seen from the foregoing brief synopsis how the climatic conditions over 
South America vary according to the position of the southern "centres of action." It 
is not sufficient, however, to confine our investigation to the Southern Hemisphere, 
because there is an intimate relation between both hemispheres. For example, the 
cyclonic " centres of action " situated in the Weddell Sea and in the Icelandic region 
stand at certain periods of the year in a definite relation to each other. In the month 
of May, in which that relation appears to be most pronounced, a positive departure 
from the normal barometric pressure at Stykkisholm (Iceland) is a direct index of a 
negative deviation from the normal in the pressure over a region to the south-east of 
and including Cape Horn. When there is a positive deviation over this latter area, 
then the departures in the Icelandic region are negative. This, be it understood, applies 
only to the month of May, and is based on twenty years' data from Cape Horn and six 
years' data from the South Orkneys ; but the whole inquiry is in an early stage of 
development, and further interesting contrasts are to be expected from a discussion of 
bipolar conditions. 



Table I. 



Percentage Frequency of Wind at Point Galera, Lat. 40° 01' S., Long. 73° 44' W. 
7 Years 1899-1905. Hours of Observation 8-2-9. 



N. 


Jan. 


Feb. 


Mar. 


April. 


May. 


June. 


July. 


Aug. 


Sept. 


Oct. 


Nov. 


Dec. 


Year. 


20 


20 


30 


42 


47 


45 


32 


33 


28 


13 


19 


15 


29 


N.E. 


1 


1 


1 


3 


5 


5 


4 


3 


2 


1 


1 





2 


E. 





1 


2 


1 


7 


6 


6 


5 


3 


1 





1 


3 


S.E. 


1 


2 


3 


5 


2 


2 


6 


5 


3 


3 


2 


2 


3 


S. 


52 


51 


36 


18 


10 


9 


9 


9 


20 


34 


32 


52 


28 


S.W. 


11 


9 


9 


6 


3 


5 


7 


8 


8 


16 


li 


10 


8 


W. 


5 


2 


4 


4 


4 


4 


9 


8 


11 


10 


6 


6 


6 


N.W. 


5 


4 


4 


10 


15 


16 


22 


22 


20 


15 


21 


7 


13 


Calm. 


5 


10 


11 


11 


7 


8 


5' 


7 


5 


7 


8 


7 


8 



Table II. 

Percentage Frequency of Wind at Ancud, Lat. 41° 51' S., Long. 73° 50' W. 6 Years 1900-05. 

Hours of Observation 8-2-9. 



N. 


Jan. 


Feb. 


Mar. 


April. 


May. 


June. 


July. 


Aug. 


Sept. 


Oct. 


Nov. 


Dec. 

11 


Year. 


9 


10 


16 


15 


11 


14 


10 


13 


18 


11 


11 


12 


N.E. 


1 


3 


2 


7 


12 


14 


13 


9 


5 


4 


3 


4 


6 


I. 


2 


1 


4 


4 


8 


8 


5 


3 


3 


2 


2 


2 


4 




3 


4 


5 


5 


15 


11 


11 


6 


6 


4 


3 


2 


6 


8. 


9 


13 


12 


15 


12 


13 


11 


12 


9 


12 


9 


8 


11 


8. W. 


10 


9 


8 


6 


4 


6 


5 


6 


5 


8 


7 


10 


7 


W. 


25 


24 


18 


12 


8 


10 


13 


16 


19 


28 


25 


24 


19 


N.W. 


22 


18 


17 


17 


16 


15 


14 


24 


24 


20 


27 


19 


20 


Calm. 


19 


18 


18 


19 


14 


9 


18 


11 


11 


11 


13 


20 


15 



THE MONSOONS OF THE CHILIAN LITTORAL. 



141 



Table III. 

Percentage Frequency of Wind at Chilian Stations fok the Months of January and June. 

5 Years 1901-05. Hours 8-2-9. 

Note. — For Angeles, owing to the. large number of calms at 8 a.m. and 9 p.m., the data refer to 2 p.m. only. For Santa 
Maria the means are for the 4 years 1900-03. 



Station. 


Lat. Long. 

S. : W. 








January. 
















June. 








N. 


N.E. 


E. 


S.E. 


S. 


S.W. 


W. 


N.W. 


Calm. 


N. 


N.E. 


E. 


S.E. 


S. 


S.W. 


W. 


N.W. 


C;ilm. 


Caldera . . . . 
Point Tortuga . 
Point Angeles . 
Juan Fernandez 
Point Carranza 
Santa Maria 
Mocha Island . 
Point Galera 
Ancud .... 
Evangelists Island 


27 03 
29 56 
33 01 
33 37 

35 56 

36 59 
38 21 

40 01 

41 51 
52 24 


. , 

70 53 

71 21 

71 38 
78 50 

72 38 

73 52 
73 58 
73 44 
73 50 
75 06 


6 

11 
2 


12 
9 

11 

18 
8 
8 


3 
3 

1 





1 
1 

1 




1 

2 
2 





2 



2 

4 


31 

1 

10 
2 
3 
1 


13 
34 
66 
62 
54 
57 
9 
5 


59 
26 
52 

26 
2 
2 

10 
9 

11 

13 


5 

24 

2 

4 

1 

2 

2 

3 

25 

26 


5 
6 
2 
1 
1 
4 
5 
5 
21 
41 


6 
22 
26 

4 
18 
20 

7 

5 
20 

6 


16 
9 
17 
2 
41 
35 
38 
44 
16 
11 


6 

8 
6 


8 

13 
6 

13 
4 


9 
22 

5 
1 
3 
1 


7 
5 
5 


8 
7 
1 
12 
7 
1 
6 
3 
8 
6 


17 

1 

5 

19 

14 

14 

11 

9 

14 

17 


23 

14 

15 

32 

4 

3 

9 

6 

6 

16 


2 
14 
2 
23 
1 
6 
4 
5 
11 
12 


7 
6 
2 
8 
12 
13 
14 
10 
18 
24 


12 

19 

47 

3 

18 

19 

5 

4 

9 

5 



Table IV. 
Mean Monthly Rainfall, Coast of Chili. 



Station. 



Caldera . 
Chanaral . 
Point Tortuga 
Point Angeles 
Juan Fernandez 

Point Carranza 

Santa Maria Island 
Point Lavapie 
Isle Mocha (West) 
Point Galera . 
Port Ancud 
Evangelists Island 



Lat. 
S. 


Long. 
W. 


Years. 


27 03 


70 53 


n 


29 00 


71 36 


6 


29 50 


71 21 


8 


33 01 


71 38 


8 


33 37 


78 50 


5 


35 56 


72 38 


* \ 


36 59 


73 52 


4£ 


37 08 


73 36 


3 


38 21 


73 58 


8 


40 01 


73 44 


8 


41 51 


73 50 


7* 


52 24 


75 06 


8 



Period. 



1899-1906 
1899-1904 
1899-1906 
1899-1900 
1902-1906 

1902 
1904-1900 
1899-1903 
1904-19.16 
1899-1906 
1899-1906 
1899-1906 
1899-19U6 



Jan. 



Ins. 

0-02 
0-00 

o-oo 

0-00 
0-72 

[0-03 

0-21 
2 48 
0-77 
2-85 
3-01 
1315 



Feb. 



Ins. 

o-oo 

0-00 
0-00 
0-02 
0-86 
013 

0-49 
343 
1-23 

4'36 
4-01 
881 



Mar. 



Ins. 

o-oo 

00 
0-15 
0-68 
2-25 

2-37 

1-56 

3-08 ! 
3-43 

6-29 

7-04 

12 03 



April. 



Ins. 
0-00 
0-06 
0-03 
0-19 
3 11 

2-12 

2-25 
2-57 
4-87 

10-35 
9-73 

1018 



May. 


June. 


July. 


Aug. 
Ins. 


Sept. 
Ins. 


Oct. 


Nov. 


Dec. 
Ins. 


Year. 


Ins. 


Ins. 


Ins. 


Ins. 


Ins. 


] 
Ins. 


0-04 


0-05 


044 


0-19 


0-03 


0-02 


0-02 


0-02 


0-83 


0-34 


0-86 


1-23 


0-82 


002 


0-04 


0-00 


o-oo 


337 


1-79 


228 


2 '22 


1-38 


0-06 


o-oi 


0-01 


o-oi 


7 94 


5-72 


6-61 


7-76 


4-68 


0-44 


0-37 


o-oo 


0-35 


23-82 


7-64 


7-01 


7-79 


334 


2-45 


1-S9 


1-40 


0-38 


38-34 


571 


9-07 


9-93 


3-30 


1-08 


1-44 


0-47 


1-07 


37-32 


6-55 


10-68 


5-87 


3-50 


1-78 


0-98 


0-43 


0-50 


34-80 


7-42 


7-52 


7 83 


2-83 


3-39 


1-43 


0-48 


0-79 


4385 


9 07 


1093 


10-08 


5-68 


3-61 


1-93 


1-33 


1-13 


54-06 


14-93 


1854 


16-28 


12-17 


8-44 


4-41 


3-52 


3-10 


105-24 


1136 


11-11 


9-76 


£-84 


8-63 


4-45 


4-09 


2-92 


84-95 


7-28 


7-72 


7-00 


7-27 


765 


10-14 


9-38 


10-83 


111-44 



Table V. 
Rainfall on Coast of Chili, July 1903 and 1904. 



July 1903 
July 1904 


Caldera. 


Chanaral. 


Tortuga. 


Angeles 


Juan 
Fernandez. 


Carranza. 


Isle Mocha. 
(West). 


Point 
Galera. 

Ins. 

8-66 
25-18 


Port 
Ancud. 


Evangelists 
Island. 


Ins. 

o-oo 

0-62 


Ins. 
073 
0-62 


Ins. 
1-74 
5-27 


Ins. 
2-57 
9-91 


Ins. 
1-62 

12-17 


Ins. 

1-02 

16-22 


Ius. 
4-66 
19-89 


Ins. 

8-70 

14-64 


Ins. 

12-81 

3-98 



TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 6). 



21 



( 143 ) 



VII. — A Carboniferous Fauna from Nowaja Semlja, collected by Dr W. S. Bruce. 
By G. W. Lee, D.Sc, H.M. Geological Survey. With Notes on the Corals by 
R. Gk Carruthers. Communicated by Dr Horne, F.R.S. (With 2 Plates.) 

(MS. received May 29, 1909. Read July 5, 1909. Issued separately September 24, 1909.) 

Introduction. 

The present paper is based on the study of a suite of fossils collected in 1898 by 
Dr W. S. Bruce, during a cruise with Major Andrew Coats in the Barents Sea, on 
board the yacht Blencathra. 

The fossils were found in situ in a cliff at the extremity of Cape Cherney, a promon- 
tory situated on the west coast of Southern Nowaja Semlja, in lat. 70° 49' and long. 
56° 37'. Contrary to what often obtains in the case of materials collected by explorers, 
they are all from the same bed, a fact which greatly increases the value of the collection, 
as there is thus no possibility of a mixing of forms from different horizons. 

The rock is a compact, brownish-grey limestone, emitting a pronounced fetid odour 
when struck with a hammer. Numerous thin sections and the residue left after attack 
by acid show that clastic minerals are small and poorly represented, and it is interesting 
to note that small patches of violet fluorite, of secondary origin, occasionally fill in 
cavities in the rock, or replace the aragonite of some of the shells. 

The limestone is essentially foraminiferal, with thin crinoidal layers, and Dr Bruce 
tells me that in the field its characteristic feature is the presence of a very large shell 
[Productus giganteus (Mart.)] ; at one spot in that same bed he also observed a large 
branching body, which may possibly represent a drifted stigmarian root. 

Most of the fossils are in a perfect state of preservation, but many suffered during 
the process of developing from the matrix, owing to their small size. In nearly every 
case the original shell-substance is preserved, calcite pseudomorphs occurring only in 
some of the Gasteropods; but even in this case the minutest details of ornamentation are 
reproduced. I wish here to emphasise the fact that all the fossils must be regarded as 
contemporaneous, there being no evidence of any having been " remanie." 

An unusual interest attaches to the fauna collected by Dr Bruce, as it proves the 
hitherto unrecorded presence of marine Lower Carboniferous rocks in Nowaja Semlja ; 
that is to say, we know now for the first time that the Lower Carboniferous sea did extend 
to such high latitudes in the western hemisphere ; this discovery has therefore more 
than a merely local importance, as it entails considerable modifications in our theoretical 
conceptions of the Arctic Lower Carboniferous Continent. For the benefit of those who 
may not be conversant with Arctic geology, I will add that geologists who have a special 
knowledge of the Upper Palaeozoic formations of these regions have pointed out in 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 7). 22 



141 DR G. W. LEE ON 

various publications that the so-called "Mountain Limestone" of Spitsbergen, Beeren 
Eiland, Nowaja Semlja, etc., really belongs to the upper divisions of the system, and 
was identified as Mountain Limestone owing to the fact that the older palaeontologists 
did not realise the importance of a closer comparison with the marine phases of the 
Upper Paleozoic formations.* 

As regards former records of Carboniferous rocks in Nowaja Semlja, very little need 
be said here. Professor Tschernyschew has ably summarised in his great work (loc. 
cit., pp. 679-681) all the data bearing on this question, the main facts being as follows: — 
According to him, the numerous species described by Professor Toula from the Barents 
[sland,t and generally believed to be Upper Carboniferous, belong to beds of Artinsk 
age ; the other Carboniferous fossils recorded from various parts of the two main islands 
are too scanty to permit of determining their true horizons. However, a few fossils 
collected during the expedition of the lermak in 1901 (Kreuz Bay and Maschigin 
Bay) indicate a horizon which is not more recent than the Middle Carboniferous 
of the Urals. 

These circumstances explain the absence of any reference by me in the descriptions 
below to species previously recorded from Nowaja Semlja, careful comparison having 
shown that this fauna is more closely related to that of the typical Lower Carboniferous 
districts of Russia and Western Europe. On the other hand, a region of which so little 
is known might have been the seat of the persistence of older types, or of the early 
appearance of forms known elsewhere from higher horizons only, so that in the determina- 
tions I have avoided confining myself to literature dealing exclusively with Lower 
Carboniferous palaeontology. 

Before passing to the descriptive part of this paper, I wish to thank Dr Bruce for 
having kindly allowed me to study his valuable collection ; and I owe a debt of gratitude 
to Professor Th. Tschernyschew, who generously furnished me with valuable information 
touching the geology of Arctic Russia. For help of a varied nature I must express my 
indebtedness to many of my colleagues, specially to Mr R. Lunn, who sacrificed much of 
his time in the preparation of the accompanying plates. 

Figs. 42, 43, 44, 46, 47, 48, and 50« are from drawings by Mr H. S. M'Vey, and 
figs. 12a, 13d, 15a, 17a, 22-22d, 246, 34a, 35, 36, 37a, and 49 are from drawings 
by myself. 

Descriptive Part. 

As the main object of the following descriptions is to bring out the faunistic and 
stratigraphical relations of the species, I have not dwelt at length on morphological 
features unless these have appeared to me to be new or of importance in systematics. 

* Frkch, Lethxa Geognostica (passim) ; J. G. Andeksson, Uber die Stiatigraphie und TeJctonik der Baren Insel, 
1899, p. 16 ; Th. Tschernyschew, " Die obercarbonischen Bracbiopoden des Ural und des Timan," Mem. Com. Gdol. 
Russie, 1902, p. 694. 

t Sitzungsb. k Ahad. Wiss. IVien, Bd. Ixxi., 1. Abth., 1875. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 145 

For the same reason I have refrained from giving synonymic lists, the references I give 
being sufficient to show the interpretation adopted in any ambiguous cases, and where 
no reference is given it is implied that the original specific designation is accepted. 

Class RHIZOPODA. 

Order Foraminifrra. 

These organisms are very abundant and well preserved ; many of them I succeeded 
in freeing from the matrix, in order to base the determinations on both external and 
internal characters. 

The Carboniferous Foraminifera of Russia have been so accurately described in the 
exhaustive monograph of V. von Moller that the task of identifying the Nowaja 
Semlja forms was a specially easy one ; consequently I have nothing of importance to 
add to what has been stated by the Russian palaeontologist.* 

Genus Endothyra Phillips. 

Endothyra globula (d'Eichwald). 
Moller, loc. tit., 1878, p. 98, pi. xiii. figs. 1-4. 

This species is represented by a number of specimens in various stages of growth ; 
one of the largest measures approximately 15 mm. in diameter and 0"8 mm. in 
thickness. 

Genus Spirillina Ehrenberg. 

Spirillina 1 sp. indet. 

Two of the slides contain horizontal sections of a non-septate tube coiled on itself in 
one plane ; as the specimens consist of seven whorls — although they measure only 
0*5 mm. in diameter — comparison with the species referred to Spirillina by Moller is 
difficult. 

Likewise I am unable to express an opinion as to their true affinities. In his review 
of the genus, H. B. Brady t states that SpirilUnse are not known to occur in formations 
older than those of the Miocene period, while making no reference to Moller's recogni- 
tion of that genus in Carboniferous strata. 

Genus Cribrostomum Moller. 

I have not sufficient knowledge of Foraminifera to form an independent opinion re- 
garding the respective claims of d'Orbigny's Bigenerina and of Moller's Cribrostomum. 

* V. ton Moller, "Die spiral-gewundenen Foraminiferen des russischen Kohlenkalks," Mem. Acad. Imp. 
Sciences de St Pe'tersbourg, 1878 ; and " Die Foraminiferen des russischen Kohlenkalks," ibid., 1879. 
+ Challenger Reports, vol. ix., 1884, p. 629. 



14G DK G. W. LEE ON 

II. 1). Brady* infers that the Carboniferous species placed under Cribrostomum by 
Moller belong to Textulana and Bigenerina, and the latter name is adopted by 
Dr H. Yabe t for Cribrostomum bradiji Moller. 

" Cribrostomum " gracile Moller. 

Moller, loc. ait., 1879, pp. 59, 60, fig. 23, pi. iii. fig. 4. 

The examples at hand appear quite typical, but I have not observed any showing 
the uniserial mode of growth, and none attain the maximum size of the Russian 
specimens. 

Genus Tetrataxis Ehrenberg. 

Tetrataxis gibba Moller. 

Moller, loc. cit., 1879, p. 73, pi. ii. fig. 4, pi. vii. fig. 3. 

As regards shape and size the few examples observed agree in every respect with the 
original definition of the species. 

Genus Arch^ediscus Brady. 

Archsediscus karreri Brady. 
Brady, Monograph of Carboniferous and Permian Foraminifera, Pal. Soc, 1876, p. 142, pi. xi. figs. 1-6. 

This species is very abundant, and the specimens average 1 mm. in diameter, being 
thus appreciably larger than the Russian examples known to Moller. 

Genus Fusulinella Moller. 

Fusulinella struvei Moller. 

Moller, loc. cit., 1879, p. 22, pi. ii. fig. 1, pi. v. fig. 4. 

Fusulinella struvei is represented by numerous specimens in various stages of 
growth, agreeing remarkably well with the Russian form as delineated by Moller. 

Genus Lagena Walker. 
Lagena ? sp. indet. 

Small flask-shaped bodies may possibly be referable to some species of Lagena, but 
as they were observed in thin sections only, they had better be left unclassified, even 
generically. 

* Loc. cit, p. 364. 

+ Journal of the College of Science, Imperial University of Tokyo, vol. xxi., article 5, 1906. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 147 



Order Radiolaria. 

Pounded portions of the limestone, after treatment with acid, indicated the presence 
of a few siliceous organisms, among which were observed unrecognisable fragments 
showing a regular lattice-work, and a whole radiolarian. The latter was filled with 
some carbonaceous or ferrugineous substance, but the lattice-work could still be 
detected, though not very distinctly. In thin sections the small bodies referred here 
to Radiolaria are more easily seen, and the following genera appear to be represented : — 

Genus Cenospelera Ehrenberg. 
Cenosphsera " sp." a. 

Simple spheres, the lattice- work of which is too indistinct for description, probably 
belong here. They vary in diameter from 0*1 mm. to 0'3 mm., but it is not 
possible to say how many " species " are present. 

Cenosphseral sp. b. 

I am in doubt concerning the generic affinities of a few small bodies, 0*1 mm. in 
diameter, consisting of an apparently simple sphere, the surface of which is covered 
with numerous short bristles. 

Genus Carposph^ra Haeckel. 

The genus is perhaps represented by small bodies measuring up to 0'2 mm. in 
diameter, and apparently made up of two concentric spheres. The distance between the 
two spheres is greater than the diameter of the smaller one ; there are still a few indica- 
tions of lattice-structure, but the connecting rays are not preserved. 

Genus Cenellipsis Haeckel. 

In the form which I doubtfully refer to this genus, the test is simple and oval, and 
its long diameter measures 0"15 mm. 

Class SPONGLE. 

The few spicules observed are slightly arcuate, with obtusely rounded extremities, 
and bear a general resemblance to those of Renieria clavata Hinde, as described by 
Dr G. J. Hinde.* The largest one is approximately - 6 long and 0"06 mm. thick. 

* A Monograph of the British Fossil Sponges, Pal. Soc, 1887, p. 143, pi. ix. fig. 5. 



148 DR G. W. LEE ON 



Class ANTHOZOA. 

Notes on the Carboniferous Corals collected in Nowaja Semlja 
by Dr W. S. Bruce. By R. G. Carruthers. 

Dr Lee has kindly allowed me to examine the corals included in the collection 
of palaeozoic fossils collected in Nowaja Semlja by Dr W. S. Bruce. For the 
opportunity thus afforded of studying an assemblage of unusual interest, I am 
sincerely grateful. 

These corals are all embedded in limestone, together with the rich fauna described 
by Dr Lee. The specimens are fragmentary, but are otherwise in a fine state of 
preservation. 

Notwithstanding the paucity of the material, five distinct genera and seven species 
can be recognised. These are : — 

Aulophyllum aff. moseleyianum (Thorn.). 
Campophyllum carinatum sp. no v. 
Campophyllum kungurense (Stuck.). 
Carcinophyllum sp. 
Lophophyllum cherneyense sp. nov. 
Lophophyllum sp. 
Hexaphyllia m'coyi (Duncan). 

All of the above genera occur in Carboniferous strata, and, with the exception of 
Campophyllum, attain their maximum development in strata of Upper Visean age, or 
in the immediately succeeding beds (Yoredale Rocks of England, Lower Limestone 
scries of Scotland, or lower Mosque nsis zone of Russia). The genera Aulophyllum, 
Carcinophyllum, and Hexapliyllia are particularly distinctive, and it is also from beds 
belonging to this general horizon that the species nearest to the Nowaja Semlja forms 
have been described. 

The whole assemblage is clearly indicative of some horizon between the base of the 
Upper Carboniferous and top of the Lower Carboniferous of the Russian classification. 
This conclusion is in harmony with that arrived at by Dr Lee from his examination of 
the other groups of fossils. 

Contrary to expectation, only one species can be definitely identified with a Russian 
form, but it is quite possible that a more complete collection might show a closer 
affinity to the Russian fauna than the present one does. 

Attention may be drawn to the revised definition of the genus Lophophyllum 
that is here furnished, while the discovery of a Carboniferous coral, Campophyllum 
ca/rinatum, with well-developed carime on the septa is certainly a point of 
interest. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 149 



Genus Aulophyllum M. Ed. and H. 

The characters relied upon by Duncan and Thomson in separating their genus 
Cyclophyllum from the older genus Aulophyllum certainly do not seem of generic value, 
and, in the writer's opinion, Cyclophyllum should be discarded. 

Aulophyllum aff. moseleyianum Thorn. 

The single specimen found of this genus was unfortunately broken off through the 
floor of the calyx, and as the lower part of the coral is missing, specific determination is 
impossible. The broken fragment, however, is identical in every way with the 
transverse section of A. moseleyianum figured by Thomson,* except that the vesicles 
of the central area seem somewhat larger than in Thomson's species. 

Genus Campophyllum M. Ed. and H. 

MM. Milne Edwards and Haime selected as the type of their genus Campophyllum 
a coral identified by them with the Cyathophyllum jiexuosum of Goldfuss; but according 
to ScHLUTERt this identification is erroneous, and their figured specimens (Polypes 
Fossiles, pi. viii. figs. 4 and 4a) belong to a new species. Such a conclusion, however, 
does not affect the validity of the genus. 

A recent examination of the genotype, for which I am indebted to the courtesy of 
the authorities at the Musee d'Histoire Naturelle, at Paris, and to M. Gravier in 
particular, shows that the original figures are of a very faithful nature. There is little 
to add to the description given in the Polypes Fossiles. It may be mentioned, how- 
ever, that a thin section cut from one of the figured specimens (fig. 4a) showed that at 
an early stage of growth the septa are amplexoid and dissepiments are developed ; the 
septa are without carinee, and the cardinal fossnla is barely discernible. 

The genus is intermediate in character between Cyathophyllum and Caninia, to 
both of which it is closely allied. In their mature growth-stages, indeed, it is impossible 
to distinguish Caninia from Campophyllum ; it has been held that the depth of the 
cardinal fossula is greater in the former genus, but this factor is so variable that little 
reliance can be placed on it. The greatest difference between these two genera is shown 
in the young growth-stages. At this period, in Caninia, the septa reach to the centre 
of the corallum, and dissepiments have not yet appeared ; whereas in a Campophyllum, 
from a very early stage onwards, the septa are short and amplexoid, and dissepiments 
apparent. In Cyathophyllum dissepiments also appear at an early stage, but the septa 
reach to the centre of the corallum throughout. 

These distinctions are not, perhaps, of a very striking character, and it may 

* James Thomson, "A New Family of Rugose Corals," Proc. Phil. Soc. Glasg., 1882, p. 33, pi. iii. fig. 12. 
t "Anthoz. d. Rhein. Mittel Devon," Abhand. Geol. Specialkarte v. Preussen, Bd. viii., Heft 4, p. 42 (1889). 



150 DR G - w - LEE 0N 

be questioned whether Caninia and Campoplujllum are in reality generically 

separable. 

The retention of the latter genus, however, seems advisable on the whole, if only for 
convenience in working purposes. 

Campophyllum carinatum sp. nov. (PL I. figs. 3-6.) 



in 



Two complete specimens and six fragments referable to this species were found 
the collection. 

EXTERNAL CHARACTERS. 

Corallum elongated, more or less cylindrical, and often slightly twisted. Largest 
diameter known, 2 cm. ; but most of the specimens are much smaller. Epitheca, thin, 
smooth, with abundant faint lines and constrictions of growth. In the young and 
conical part of the corallum, faint longitudinal ribbing, with the pinnate arrangement 
characteristic of Rugose corals, can sometimes be seen. Calyx unknown except in 
vertical section. (PI. I. fig. 3.) 

INTERNAL CHARACTERS. 

(a) Transverse Sections. — Major septa of uniform thinness throughout, closely 
set, and never reaching to the centre of the section. The base tabular area in the 
centre of the corallum is of very variable width (compare figs, da and 4, PI. I.), 
owing to the amplexoid nature of the septa. The minor septa are long, and 
project a short distance past the dissepiments into the tabular zone. Both major 
and minor septa possess well-marked carinas, appearing in transverse sections as 
minute irregular teeth-like projections along the septa. No distinct fossula has been 
observed. 

(b) Vertical Sections. — The tabula? are numerous, horizontal in the centre of the 
corallum, but curving abruptly downwards before reaching the dissepiments, which are 
small, globose, and developed at an early stage. The breadth of the tabulate zone is 
uniform in the more mature part of the corallum, but that of the dissepimeutal zone is 
very variable (see figs. 3 and 5, PI. I.). 

When the vertical section is cut down the centre of the corallum and passes down 
the surface of a septum, the true nature of the carinas can be readily observed ; they 
form closely set spiny ridges running upwards and inwards over the surface of the septa, 
in a somewhat irregular manner (see figs. 3 and 5, PI. I.). Under a high magnifica- 
tion, in the centre of certain of these ridges there is a line of minute black points from 
which radiating fibres proceed. These black points are probably centres of calcification, 
and indicate that the carinal ridges are composite, being formed by the fusion of closely 
set spines, lineally arranged. 

Remarks. — There are many points of resemblance between this interesting species 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 151 

and Thomson's Campophyllum echinatum* but in the latter there are no carinas on 
the septa, and the tabulae are much flatter. The same distinctions also separate Milne 
Edwards and Haime's genotype from the new species. A deceptive resemblance to the 
inner wall of the Devonian genus Craspedophyllum is often seen in transverse sections, 
owing to the intersection of a tabula against which the septa terminate, or in vertical 
sections from the intersection of two septal edges within the tabulate area. The great 
interest of this new species lies in the carinse that cover the septal surfaces. Such a 
phenomenon is, of course, common in Devonian corals, but is of extreme rarity in 
Carboniferous species. A low stratigraphical position for the species does not necessarily 
follow. Cyathaxonia rushiana Vaughan, which seems to be the only Carboniferous 
coral yet found to have carinse, comes from horizons very high in, or above, the Visean. 
[An examination of C rushiana shows that the projections on the septa seen in trans- 
verse sections, to which Dr Vaughan has drawn attention (Quart. Jour. Geol. Soc., 
1908, p. 460), are in reality carinse.] 

Campophyllum kungurense (Stuckenberg). 

Amplexus Tcungurensis, A. Stuckenberg, "Korallen u. Bryoz. d. Ural u. d. Timan," Mem. Com. 
Geol. Eussie, vol. x. No. 3 (1895), p. 189 of German text, pi. iii. figs. 10-10c. 

Only one example of this species has been found in the collection, but fortunately 
the specimen is complete and well preserved. 

All the characters are in perfect agreement with Stuckenberg's figures and descrip- 
tions, and there is little to add to his diagnosis. The septa become amplexoid at a very 
early growth-stage, and a vertical section down the calyx confirmed Stuckenberg's 
observation that in the nature part of the corallum a single row of rather large dissepi- 
ments appears between the tabulae and the wall. Such a feature is not found in a true 
Amplexus (e.g. A. coralloides), and the coral may more fitly be placed in the genus 
Campophyllum. 

Campophyllum kungurense is practically identical with the common spineless form 
of the Tournaisian coral Campophyllum spinosum (Amplexus spinosus de Kon.). An 
examination of several specimens of the latter species, procured from Tournai, showed 
that the septa and dissepiments were somewhat thinner than in C. kungurense, but no 
other difference was discernible. It is probable that such a distinction is not of specific 
value, but, pending further investigation, Stuckenberg's name has been retained. His 
specimens were obtained from the east side of the Urals, presumably from strata of 
Permo-Carboniferous age ; it seems probable, therefore, that we are dealing with a 
remarkably stable type, ranging throughout Carboniferous time with but little structural 
alteration. 

* James Thomson, "A New Family of Rugose Corals," Proc. Phil. Soc. Glasgoiv, 1882, pi. iv. figs. 10, 10a. In 
his large paper, published by the Society in the succeeding year, the same figures are referred to Campophyllum 
paracida M'Coy. 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 7). 23 



152 DR G. W. LEE ON 

Genus Carcinophyllum Thomson. 

Carcinophyllum sp. (PL I. figs. 8-8&.) 

Besides the figured specimen, another example, complete but of small size, was 
obtained. 

The corallum is conical and very slightly curved. The calyx is deep and the 
epitheca smooth, with slight accretions of growth. In the mature growth-stages, when 
it is separated from the septa, the epitheca is seen to be of medium thickness, but at 
the proximal end of the corallum appears to be of great thickness, owing to the complete 
fusion of the septal bases. 

A comparison of figs. 8 and 8b shows that, with increasing age, the complexity of 
the lamellae in the central area increases, an Aulophyllum-Yike character being assumed. 
Such a fact, observable also in other species of the genus, suggests that the Aulophylla 
may have been evolved directly from the Carcinophylla. It is interesting, therefore, 
to note that Dr Vaughan finds the maximum of the genus Carcinophyllum to be at 
the base of his D x sub-zone, while the Aulophylla appear much later, at the top of the 
D 2 sub-zone. 

Such a genetic connection as is here suggested does not, of course, preclude the 
possibility of representatives of these two genera existing side by side. They un- 
doubtedly do so in this Nowaja Semljan fauna, and also in the Scottish Carboniferous 
Limestones ; Dr Sibly has also found such an association in the D 2 sub-zone of the 
English Midlands. 



*&' 



Genus Lophophyllum M. Edw. and H. 

1845. Cyathaxonia (pars), Michelin. 

1850. Lophophyllum, M. Ed. and H. 

1876. Koninckophyllum, Nicli. and Tliom. 

1883. Acrophyllum, Thorn, (non Nicli.) ; non Lophophyllum, Nich. and Thorn. 

The opportunity is here taken to give a revised diagnosis of the genus Lophophyllum,, 
two species of which occur in Dr Brtjce's collection. 

Michelin, in 1846, when describing the corals of Tournai, referred to his genus 
Cyathaxonia a new species which he named C. tortuosa. Later, in 1850, Milne 
Edwards and Haimr established the genus Lophophyllum, selecting as the type 
L. Jconincki, which also came from Tournai. 

An examination of a large number of topotypes of these two "species" shows that 

(1) Lophophyllum Jconincki is simply a young form of " Cyathaxonia" tortuosa, which 

(2) cannot be referred to the genus Cyathaxonia, as in the mature growth-stages a 
well-marked zone of dissepiments appears between the tabulae and the wall, and such a 
feature is quite absent in Cyathaxonia proper (e.g. C comu Mich., or C. rushiana 
Vaughan). 

The genotype of Lophophyllum therefore is L, tortuosum (Mich.), and the revised 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 153 

diagnosis of the genus, founded on an examination of this and allied species, is appended 
below : — 

Corallum simple and turbinate. Major septa meeting in the centre of the 
coral in the early growth-stages. One of the septa, usually the counter septum, is 
strongly thickened at the inner end, giving rise to a prominent columella, which may be 
discontinuous. In the more mature growth-stages, the columella persists, but the septa 
usually retreat from the centre and become amplexoid, while dissepiments appear 
between the tabulae and the wall. The tabulae are arched upwards in the centre to 
a varying degree, but, unlike such genera as Dibunophyllum, there is no central zone 
where the tabulae are more numerous or vesicular, nor is there a system of vertical 
lamellae distinct from the septa. 

All the simple forms of the genus Koninckophyllum of Nicholson and Thomson, 
and also those corals referred by Thomson to Acrophyllum* thus fall into Lophophyllum 
as above denned. It is suggested that the compound forms of Koninckophyllum should 
also be included, as they only differ from a true Lophophyllum, in their pronounced 
lateral budding. From Lithostrotions of the L. martini type the genus is distinguished 
by being almost always simple and not forming tubular colonies. 

Those corals referred by Nicholson and Thomson and other authors to Lophophyllum 
(e.g. L. proliferum and L. eruca) do not develop dissepiments at any stage of growth, 
and are essentially Zaphrentes having one of the septa thickened at the inner end. It 
may be convenient at some future time to group them as a sub-genus of Zaphrentis, 
but for the present such a course is not considered advisable. 

Lophophyllum cherneyense sp. nov. (PI. I. figs. l-2d.) 

EXTERNAL CHARACTERS. 

Corallum cylindrical and twisted for the greater part of its length ; conical and 
curved only at the proximal end. Epitheca thin, smooth, with numerous constrictions 
of growth, but no longitudinal ribbing. Calyx unknown. 

The largest specimen measures 2'5 cm. by 4 cm., but is incomplete; eight other 
fragments were also found. The full length may have been 8 cm. or more. 

INTERNAL CHARACTERS. 

Transverse Sections. — Major septa thin, straight, and tapering towards the inner 

end ; no stereoplasmic thickening. In proximal sections (PL I. figs. 2c, 2t/) they all 

reach the central columella formed by a thickening of the inner ends of the cardinal and 

counter septa. In later growth-stages, most of the septa fall away slightly from the 

columella, although a few are still prolonged over the central area in an irregular 

manner. 

* Thomson was mistaken in his conception of this genus, which was distinctly stated by Nicholson to have no 
columella. Good figures of the genotype, A. oneidense (Billings) are given by Lambe in Contrib. lo Canadian Pal., 
vol. iv. part 1, pi. xvi. figs. 1 and 2 (1899). 



154 DR G. W. LEE ON 

Minor septa fairly well developed, but irregular and not reaching to the tabular area. 
The cardinal fossula is indistinct, and can only be recognised in some sections by a 
shortening of the cardinal septum, towards which the axial columella is directed. 

Vertical Sections. — The median vertical section figured (PI. I. fig. 2) sufficiently 
indicates the nature of the tabulae and dissepiments. The latter are developed at an 
early growth-stage. 

lie mi irks. — Certain sections (e.g. PL I. fig. 2a) of this species closely resemble 
that figured by Thomson * as Acrophyllum sp., but in the latter the major septa are 
allccted by a thickening of stereoplasma within the tabular area. 

In distal transverse sections (PI. I. fig. 1) the central area of Lophophyllum cher- 
neyense approaches in structural character that of a Dibunophyllum. As regards the 
septa and columella, the Lophophyllwm stage is strongly marked in the younger stages 
of the new species (PL I. figs. 2a, 2d), while the tabulae throughout approximate to a 
Lophophyllum type, as opposed to Dibunophyllum, where the tabulae have a distinct 
central area in which they are more numerous and vesicular. 

Probably all the Dibunophylla pass through a Lophophyllum stage, but assume 
their characteristic features at so early a period that as a rule they are readily separable 
from the later genus. 

Lophophyllum sp. (PL I. figs. 7, 7a.) 

Of this species, apparently new, there is only a single example in the collection, and 
the conical end is missing. The corallum is continuously conical and slightly curved. 
The calyx is deep (4 cm.), the diameter of the rim being 4 - 5 cm. The ejpitheca is thin, 
smooth, with a few lines and constrictions of growth. The columella is a prominent 
feature but from the vertical section (PL I. fig. 7a) is apparently discontinuous. 
As the specimen is embedded in limestone, it is not possible to say whether this is in 
reality the case, or is only due to a slight twist in the corallum taking the columella out 
of the plane of section at this point. 

The cardinal fossula is clearly visible in cross section (PL I. fig. 7), owing to the 
sharply marked boundary of the tabular area encroaching on the dissepimental zone at 
this point. 

Incert^ Sedis. 
Genus Hexaphyllia Stuckenberg. 

Heterophjllia, M'Coy (pars). 

The genus Hexaphyllia was established by Stuckenberg t for those species of 
Heterophyllia with six septa only. Although the multi-septate forms of M'Coy's genus 
(e.g. Heterophyllia grandis) probably pass through a hexameral stage early in life, still 

* " Corals Carb. Syst. Scotland," pi. xi. figs. 14 and 14a, Proc. Phil. Soc. Glasgow, 1883. 

+ " Anthoz. u. Bryoz. d. miter. Kolilenkalks v. Central Russlande," Mem. Com,. Geol. Russie, nouv. ser., livr. xiv. 
(11MJ4), p. 72 of German text. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 155 

on the whole the separation of the hexameral species seems advisable. Many of the 
latter are of considerable size, and certainly cannot be regarded as early stages of the 
multi-septate forms. 

These peculiar genera are difficult to classify. They differ from ordinary Palaeozoic 
corals in at least two important respects. In the first case, certain species apparently 
have no epitheca, for the dark median line of the septa passes completely through the wall 
to the exterior, instead of being interrupted by an epitheca. In the second case, in 
certain species a protoseptal stage with four septa at right angles to one another seems 
to be indicated, and such an arrangement is at variance with our knowledge of other 
Anthozoa. The remarkable external ornamentation of such species as Hexaphyllia 
mirabilis, and the general habit of growth, are also peculiar. 

Further investigation of these curious genera is certainly needed before their 
systematic position can be secured. 

Hexaphyllia m'coyi (Duncan). (PL I. fig. 9.) 

Heterophyllia m'coyi, Duncan, Phil. Trans., clvii. p. 645, pi. xxxi. figs. 3a, 3c. 

Heterophyllia m'coyi, Thomson, "Corals Curb. Syst. Scotland," Proc. Phil. Soc. Glasgoiv, 1883, p. 118, 
pi. x. figs. 20, 20a. 

There are several fragments of this species, varying from 2 to 6 mm. in diameter. 

The corallurn has a smooth exterior and is six-sided, with parallel walls. There is a 
low ridge at each angle, to which one of the six septa extends. The dark median line 
of each septum passes through the wall (which is very thick) to the exterior, and in 
weathering out gives rise to a faint longitudinal groove down the middle of each of the 
external ridges (" costae " of Duncan). No complete specimens have yet been recorded ; 
the whole corallurn may have been 20 cm. or more in length, tapering with extreme 
slowness. 

In transverse sections of mature examples the septa are rather thick, and have a 
regular hexameral disposition. In young specimens they are cruciform, one arm of the 
cross being forked at each end ; while six septa are therefore still present, it appears 
possible that in still younger stages there may be only four, arranged at right angles 
to one another. 

The tabula; are very close together ; they are bent sharply downwards near the wall, 
and upwards in the centre, but are almost flat in the middle. The appearance of a 
vertical section varies according to its distance from the centre of the coral. If the 
section be cut close to the wall, the tabulae appear very steeply inclined upwards, but 
when cut down the centre of the coral they are more flattened (see PL I. fig. 9). 

Remarks. — Neither Duncan nor Thomson figures any vertical section of this species, 
the former simply remarking that " the endotheca [ = tabulae] is very abundant." 

The type was not available for examination, but another specimen, in every way 
identical with that figured by Duncan, and procured from a similar horizon in Ayrshire, 
is in the collection of Mr James Neilson of Glasgow. Mr Neilson very kindly allowed 



150 Dll G. W. LEE ON 

this specimen to be cut, so that the nature of the tabulae could be ascertained, and a 
very necessary addition made to the diagnosis of the species. 

The fragments from Nowaja Seinlja are in complete accordance with the Scottish 
specimens. 

/ /< xaphyllia m'coyi resembles H. prismatica Stuckenberg,* but in the latter the 
tabulae are much farther apart. 

Class CRINOIDEA. 

Crinoid stems are very abundant, and some of them attain a large size ; but as no 
portions of calyx were observed, any attempt at identification is out of the question. 
However, two forms at least appear to be represented. 



Class BRYOZOA. 

This class is very poorly represented ; the few examples at hand, belonging to the 
Suborders Trepostoinata and Cryptostomata, are specifically indeterminable. Yet the 
presence of a Stenopora is worth noting, as the genus is not mentioned in Stucken- 
berg's recent monograph of the Lower Carboniferous Corals and Bryozoa of Russia. t 

Genus Stenopora Lonsdale. 
Stenopora sp. 

Although thin sections could not be made, on account of paucity of materials, the 
generic position of the fragments is proved by the nature of the mature end of the 
zosecia, the casts of which exhibit swellings corresponding to unthickened portions of 
the wall. 

Fenestellid indet. (cf. Polypora papillata M'Coy). 

A small fragment of a Fenestellid, the obverse face of which is not shown, bears a 
general resemblance to Polypora papillata M'Coy. The arrangement of the fenestrules 
is the same, and, in common with it, it possesses a small papillated pore at the origin of 
most of the dissepiments, a character shared by Polypora spininodata Ulrich,} which 
is otherwise different. 

Incert^e Sedis. 

I place tentatively among the Bryozoa an indeterminable fragment possibly belong- 
ing: to the genus Cystodictya Ulrich. 

* Luc supra cit., p. "ri of < ferman text. 

t Mrm. Com,. '•<'"!. Russie, L904. 

| Ural. Surnij of Illinois, vol. viii., I»!)U, pi. lx. fig. 3. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 157 

Class BRACHIOPODA. 

Judging from the materials at hand, Brachiopods are highly characteristic of the 
Cape Cherney limestone, where they probably exhibit a rich and varied character, to 
which full justice cannot be done here, on account of the fact that many of the most 
interesting forms observed came to grief when being developed, or were so sparsely 
represented as to preclude the possibility of making satisfactory serial sections. 

Modern workers on fossil Brachiopods have shown that within certain groups there 
can occur remarkable phenomena of homoeomorphy and convergence which often make 
it impossible to estimate the true genetic affinities of many species, unless a knowledge 
of their internal characters be obtained. Such instances of forms differing in the nature 
of their internal characters, although apparently nearly allied so far as their general 
appearance indicated, were met with in a few cases, especially among the Spiriferids, 
and will be referred to in due course. 

The mode of .occurrence of the various species shows that peculiar conditions, not 
yet understood, affected the development of some forms more than others. For 
instance, it is difficult to explain how it is that Productus longispinus J. Sow., is 
represented by full-grown individuals only, as is also Productus giganteus (Mart.), 
whilst Productus elegans M'Coy, is represented by a series of specimens ranging in 
length from a few millimetres to about 40 mm. The explanation may be that conditions 
which killed the immature individuals of the latter, left those of the former unaffected ; 
this explanation seems more satisfactory than that based on the assumption that the 
nature of the shell in the young stages of Productus longispinus and Productus 
giganteus was such as to render fossilisation impossible. 

The assemblage as a whole is quite normal, and there is no evidence of dwarfing or 
of a depauperised condition. 

Terebratulid^e. 

Genus Dielasma King. 

Dielasma lenticulare de Koninck. (PL I. figs. 10-1 0c.) 
Db Koninck, Faune du Calcaire Carbonifere, 6 roe partie, 1887, p. 17, pi. 2, figs. 1-9. 

Few of the numerous species of Dielasma established by de Koninck have been 
unreservedly adopted in pal?eontological literature, but the well-defined characters of 
his Dielasma lenticulare justify its claims to specific distinctiveness. The Dielasma I 
refer to it is, like the Belgian form, lenticular and globose, with non-sinuate valves. 
Frontally, the junction line is perfectly straight. It is represented by a complete 
specimen, the dimensions of which are : length, 7 mm. ; breadth, 6 mm. ; depth, 5 mm. 
A large pedicle-valve, measuring approximately 20 ram. in length, probably belongs to 
the same species. Comparison with Davidson's figures will show that it cannot be 



158 DR G. W. LEE ON 

confounded with .any of the forms usually referred to Dielasma hastatum (J. de C. Sow.), 
and it does not appear to have closely allied representatives in the upper divisions of 
the Carboniferous system in Eastern Europe. 

Dielasma gillingense (Davidson). 

Davidson, British Carhoniferous Braeh.iopoda, Pal, Soc, 1857-1862, p. 17, pi. iii. fig. 1. 

Two small incomplete specimens, characterised by their very depressed shape, 
probably belong to this species. 

SpiRIFERIDiE. 

Genus Spiriferina d'Orbigny. 

Spiriferina insculpta (Phillips). 
Davidson, loc. cit., p. 42, pi. vii. figs. 48-55. 
The collection contains a few typical examples of this species. 

Spiriferina insculpta (Phillips), var. (PI. I. figs. 11-116.) 

Six specimens, in various stages of growth, differ in no essential degree from the 
British representatives of Spiriferina insculpta ; the only difference is that the strongly 
marked growth-lines typical of the species are here very indistinct, although the test is 
well preserved. Failing to obtain access to the internal characters, owing to the presence 
of infiltrated calcite, I am unable to say how far this surface feature should affect the 
systematic position of these specimens. 

Waagen* and S. NikitinI' have indicated the differences between Spiriferina 
insculpta and the Upper Carboniferous Spiriferina ornata Waagen. 

Spiriferina cristata (Schlotheim), var. octoplicata (J. de C. Sowerby). 
Davidson, loc. cit., p. 38, pi. vii. figs. 37-47. 

All the external characters of this common fossil are well exhibited in the materials 
at hand, consisting of a specimen nearly complete and a few fragmentary ones. The 
area is large, very much like that depicted by Davidson, fig. 37 (loc. cit.). 

Spiriferina sp. indet. 

A very imperfect specimen, exhibiting the characteristic punctate shell-structure of 
the genus Spiriferina, differs from the above in having the ribs more sharply defined, 
that is, higher and narrower. Too little is seen to enable one to say whether this 
difference is to be attributed to anything more than an individual character. 

■ Pal. Indica : Salt Range Fossils, 1883. t Mem. Com. Ge'ol. Russie, t. v., No. 5, 1890. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 159 

Genus Spirifer J. Sowerby. 

Spirifer cf. bisulcatus J. de C. Sowerby. 

The specimen is too incomplete for a satisfactory determination, but it certainly 
belongs to the group of Spirifer bisulcatus J. de C. Sowerby. 

Cf. Spirifer triradialis Phillips. 

One specimen, measuring 6*5 mm. in length and 3'5 mm. in breadth, should possibly 
be placed here, but its true affinities must remain doubtful, on account of its fragmentary 
condition. 

Genus Martiniopsis Waagen. 
Martiniopsis ? sp. (PL I. fig. 12.) 

Two pedicle-valves of a Spiriferid show a peculiar combination of characters which 
might determine its reference to either of the two genera Spirifer or Martiniopsis, 
according to the view taken as to the value of these characters, considered individually. 

The outline is sub-rhomboidal, with length exceeding width. The larger specimen 
is 18 mm. long and 16 mm. wide : it is not very convex and has gently sloping sides. 
The area appears to be small and the beak is not much incurved. The sulcus is narrow 
and does not increase in width towards the anterior margin. The surface — excepting 
the cardinal slopes, which are smooth — is ornamented by ten broad, very fiat ribs, separ- 
ated by narrow sulci. One rib occupies the bottom of the median sulcus. The external 
layer is minutely pitted, as in Martiniopsis and in most species of Martinia. Inter- 
nally, the shell is characterised by two well-developed dental plates. 

The true generic affinities of this form cannot be determined, owing to the brachial 
valve (typically septate in Martiniopsis) being unknown ; but it is interesting to note 
that the elongate shape, the median sulcus, and the ribbing of the shell are characters 
not included in Waagen's diagnosis of his genus Martiniopsis, under which the present 
form ought to be placed, on account of its pitted surface and small area. Even in the 
event of the latter view being proved correct, this would not be the first record of an 
elongate and ribbed Martiniopsis* 

Genus Martinia M'Coy. 
Martinia sp. (cf. Spirifera planata Phillips). (PI. I. fig. 14.) 

A pedicle-valve, 11 mm. in length and 10 mm. in breadth, is externally very much 
like "Spirifera" planata Phillips, as redescribed by Davidson (loc. cit., p. 26, pi. vii. 
figs. 25-36), with this difference, that the sulcus is more sharply defined and ornamented 
by three ribs only, instead of four or five. 

* Compare, for instance, Martiniopsis aschensis, Tscherntschew, loc. cit., pi. 1., fig. 4. 
TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 7). 24 



160 DR G. W. LEE ON 

Closer comparison is not possible, as the internal characters of Phillips' type- 
specimen are not known ; the species described here has no dental plates, and the surface 
is apparently not punctate. 

Martinia brucei sp. nov. (PL I. figs 13-13e.) 

This shell is depressed, sub-rectangular, with width slightly in excess of length ; the 
hinge-line is shorter than the greatest width of the shell. The sulcus and the mesial fold 
are indistinct. Both valves are very shallow and of equal depth, and the beak is not 
much incurved. If I am right in assuming that the two specimens described here under 
this name really belong to the same species, then the ornamentation varies considerably 
with age ; on the small one it consists of eighteen rounded ribs, whilst on the larger speci- 
men there are thirty such ribs, a few of which are due to bifurcation. In both, the surface 
shows an extremely minute pitting, and is covered with sharp, close-set lines of growth 
quite visible to the naked eye. The area is very narrow and there are no dental plates. 

The smaller specimen, in which both valves are in apposition, has the following 

dimensions : — 

Length . . . . .13 mm. 

Width ..... 13-5 „ 

Depth . . . . . 6 "5 „ 

The larger one is a pedicle-valve 15 mm. long and 18 mm. wide. 

I do not know any Carboniferous Martinia which could be considered as closely 
allied to this species. The nature of the ribbing and the lack of distinct sulcus and 
fold give it a general resemblance to the young stage of Martinia ? inteyricosta 
(Phillips), as figured by Davidson (loc. cit., pi. ix., figs. 17, 18), from which it differs, 
however, in having a much narrower area and the pedicle-valve not more convex than 
the brachial one. There is also a curious superficial resemblance between the smaller 
specimen and the otherwise very different Martiniopsis baschldrica Tschernyschew,* 
this being a striking example of close external resemblance between two forms which, 
according to the modern views, are to be considered as generically distinct on account of 
their internal characters. 

Martinia sp. (PL I. figs. 15-1 5a.) 

Two pedicle-valves, belonging to a species apparently undescribed, are characterised 
by their rhomboidal shape, linguiform sinus, and peculiar ornamentation consisting of 
faint ribs so flat that the narrow intervening sulci are more conspicuous than the ribs 
themselves. No trace of ribbing can be seen on the cardinal slopes. The larger 
specimen, 24 mm. long and about 20 mm. wide, has some twenty ribs ; the smaller 
one, 8 mm. long and 7 "5 mm. wide, has only a little over half this number. 

The whole surface of the shell is covered with fine longitudinal strise regularly 

* Tschehnyrchew, Inc. cit., pi. Ixiii. fig. 1. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 161 

distributed, ten of which occupy the space of one rib. No pitting is evident. There are 
no dental plates. 

This species might be compared with Martinia rhomboidalis M'Coy [non Girty, 
1908],* in common with which it has steeply falling sides, a linguiform sulcus, and 
obsolete ribbing. Its elongate shape, flatter ribs, and characteristic longitudinal striae 
distinguish it from M'Coy's species, which is also appreciably smaller. 

The peculiar striae so conspicuous here have been described in many species of 
Spiriferids from various horizons, and cannot be said to be distinctive of any particular 
group, f 

Genus Reticularia M'Coy. 
? Reticularia imbricata (J. Sowerby). 

A small and incomplete specimen has the ornamentation characteristic of this species, 
but, as I have been unable to see its internal characters, its true position must remain 
doubtful. 

Genus Squamularia Gemmelaro. 
Squamularia sp. a. (PL I. figs. 18-186.) 

The collection contains a small Spiriferid, the general outline and ornamentation of 
which would determine its reference to Reticularia lineata, were it not that it lacks all 
traces of the septum and dental plates characteristic of the genus Reticularia, as defined 
by M'Coy. Ten specimens in different stages of growth up to 12 mm. were examined, 
and all are remarkable for their extremely small and indefinite area. 

The dimensions of the specimen figured are : — 

Length ..... 10 mm. 
Width . . . . . 10 „ 

Depth . . . . 3.5 „ 

Squamularia sp. b. (PL I., figs. 19-196.) 

A few examples of a Squamularia differ from specimens of the same size belonging 
to sp. a in having a well-defined area and a more robust articulating process. Their 
ornamentation is the same as that in sp. a. The dimensions of the specimen figured 
are : — 

Length ..... 6 mm. 

Width . . . . 6 „ 

Depth . . . . . 3-5 ,, 

* M'Coy, Synopsis Garb. Limestone Fossils of Ireland, 1844, pi. xxii. fig. 11. 

t Good figures of this surface structure are given in Palseontology cf New York, vol. viii. (passim), and in Professor 
Tschernyschew's Obercarbonischen Brachiopoden (passim). 



162 DR G. W. LEE ON 

I have described these two forms separately in order to show their distinguishing 
features, but whether these have a specific value cannot be stated at present. 

Comparison with the known species of Squamularia and with many of the figures of 
" Reticularia lineata " is difficult, owing to the general likeness obtaining between the 
various members of these two groups, so that the claims to specific distinctiveness of 
the two forms described here cannot be settled without a detailed examination of 
materials from various parts of the world. 

Remarks on the Classification of the Spiriferids. 

Considerable difficulty was experienced in classifying the species enumerated above, 
owing to the internal characters of some of them pointing to generic (?) affinities different 
from those indicated by the external features. 

An elongate and ribbed Martiniopsis^.), if represented by the pedicle- valve only, is not 
easily distinguished from those forms of Spirifer which Mr S. S. Buckman has recently 
proposed to place under M 'Coy's genus Brachythyris, as it has not yet been conclusively 
proved that a small area, the presence of dental plates, and a pitted surface are characters 
not obtaining in this group. Martinia brucei sp. nov., is more strongly ribbed than 
most Martinise, and bears, in its young stage, a striking resemblance to an Upper 
Carboniferous shell referred to Martiniopsis; and Squamularia sp., is hardly distinguish- 
able from forms referred to Reticularia lineata. 

That externally similar shells may differ in the nature of certain internal features 
has long been known, but palaeontologists have not always agreed as to what degree of 
importance should be attached to these particular features, to wit, in this case, dental 
plates. Hall and Clarke* indicated these characters in some forms of their " Glabrati" 
section, but without formulating a definite opinion as to their value, and the genus 
Reticularia was believed by them to include forms devoid of dental plates, this belief 
being probably based upon Waagbn's re-definition of the genus.t But neither Waagen 
nor Hall and Clarke paid sufficient attention to the fact that M'Coy, when establishing 
his genus Reticularia, emphasised in his diagnosis the presence of a septum and dental 
lamellse,! a character observed in 1887 by Professor Tschernyschew in a specimen 
from Vise, figured for comparison with Devonian materials. § 

In 1895 Dr A. Tornquist commented upon the generic value of the term Reticularia, 
and restricted it to its original implication ; || and under the name of Martinia lineata 
M. A. Julien figured in 1895 a Reticularia in which a septum and dental plates are 
discernible.lf 

* Palaeontology of New Yorlc, vol. viii., 1893, and Introduction to the Study of the Brachiopoda, 1893. 
+ Waaobn, Pal. Indica: Salt Range Fossils, p. 538, 1887. 
t M'Coy, Synopsis, p. 142, 1844. 

§ Tschernyschew, " Die Fauna des mittleren und oberen Devon am West-Abhange des Ural," Mem. Com. Ge'ol. 
Russie, 1887, pi. x. fig. 1. 

|| A. Tornquist, Das fossilfiihrende Untercarbon am ostlichen Rossbergmassiv in den Vogesen, 1895, p. 119. 
IT A. Jolien, Le Terrain Carlonijhe marin de la France Gentrale, 1896, pi. ii. fig. 10. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 163 

Dr G. H. Girty, who treats of this matter very exhaustively, proposes to restrict 
Reticularia to its original implication, and quotes a number of American Mississipian 
forms referable to this genus. * 

English authors who have of late years given much attention to the study of 
the internal organisation of Carboniferous Brachiopods also consider dental plates 
to be characteristic of Reticularia, but no mention is made by them of the 
presence of a median septum,t and it would be rather interesting to know whether 
there are actually such forms of "Reticularia" with dental plates, but without 
median septum. 

Now, the interesting point already fully elucidated by Dr Girty is that there are 
forms having the typical ornamentation of " Spirifer lineatus" but quite devoid of 
dental and septal plates, and for these forms Dr Girty adopts Gemmelaro's genus 
Squamularia, with this interesting remark, that in America it is an Upper Carboni- 
ferous genus, whilst the septate Reticularia is essentially Lower Carboniferous. On 
the other hand, such is not the case in Europe, as proved by the researches of various 
authors. Waagen, who compared his Indian materials with specimens of " Reticu- 
laria" lineata from Vise, found the latter to be devoid of internal partitions, and 
Dr H. Scupin, after examining a number of specimens of "Spirifer" lineatus from 
Germany, stated that this species is usually without dental plates.| In this same work 
he points out that German examples of Spirifer glaber often have dental plates, and 
on this evidence he denies a generic value to Reticularia and Martinia, a point of view 
shared by E. Schellwien.§ 

The genus Squamularia having been created for the reception of Permian species, 
the above review was necessary in order to show that its presence is not incompatible 
with the Lower Carboniferous affinities of the Cape Cherney fauna, since the Lower 
Carboniferous rocks of the Continent are said to contain representatives of a non-septate 
" Spirifer lineatus," and such forms have also been recorded from Britain. || 



Genus Amboccelia Hall. 

? Ambocadia urei (Fleming). 

This species is possibly represented by an incomplete specimen of which I have not 
been able to determine the internal characters. 

* G. H. Girty, "The Carboniferous Formation and Fauna of Colorado," U.S.G.S., Professional Paper No. 15, 
1903, p. 31. 

t T. F. Sibly, " On the Carboniferous Limestone in the Mendip Area," Quart. Journ. Geol. Soc, 1906, p. 375 ; 
A. Vaughan, "On the Faunal Succession in the Carboniferous Rocks at Loughshinny," ibid., 1908, p. 469; S. S. 
Buckman, " Brachiopod Homceomorphy," ibid., 1908, p. 31. 

J H. Scupin, "Die Spiriferen Deutschlands," Palseontographica, 1900, pp. 5 and 6. 

§ E. Schellwien, " Die Fauna der Trogkofelschichten," Abh. h. k. Geol. Reichsanstalt, 1900, p. 67. 

|| A. Vaughan, "The Carboniferous Limestone Series of the Avon Gorge," Bristol Naturalists' Socitty Proceedings, 
1906, p. 159 ; S. S. Buckman, loc. cit., 1908, p. 33. 



164 DR G. W. LEE ON 

Khynchonellid^e. 

The collection contains only a few and totally indeterminable fragments of 
Rhynchonellids. 

Athyrid^e. 

Athyris ? spp. indet. (PI. I. figs. 20, 21.) 

The Athyrids are represented by a few small specimens referable to two distinct 
forms. Their shape is very depressed, but as this character appears to be partly due to 
crushing, they had better be left undetermined. 

Genus Camarospira Hall and Clarke. 

Camarospira ? sp. (PL I. figs. 22-22c£.) 

A small Brachiopod exhibiting remarkable internal features appears to be specially 
abundant in the Cape Cherney Limestone ; unfortunately, none of the specimens 
obtained is complete enough to permit of a satisfactory determination. 

Externally, the shell has an athyroid aspect, but the pedicle- valve is much larger 
than the brachial one, and has a high and apparently incurved beak. The brachial 
valve is much flatter than the pedicle-valve, and its umbo is very low. The hinge-line 
is arcuate, and the cardinal area appears to be obscure, but the shoulders are well 
marked. The apical region being broken in all the specimens, the presence or absence 
of a foramen could not be determined. 

Both valves are evenly convex but for a faint depression in the anterior portion of 
the pedicle-valve, and a certain mesial gibbosity in the brachial valve. 

The shell is quite smooth, and its structure is fibrous and impunctate. 

Internally, the pedicle-valve is characterised by highly developed and strongly 
convergent dental plates uniting to a high septum and thus forming a spondylium 
which extends to about three-fourths of the length of the shell. The supporting septum 
appears to reach the frontal margin, so that longitudinal splitting is common among 
the specimens observed. 

Of the brachial valve, two specimens only were available for study. There is no 
evidence of a spondylium ; but a septum, the dimensions of which could not be 
estimated, is present and exhibits a thin black median line. The nature of the latter is 
not easily explained, since the septum does not seem to be the result of the joining 
together of septal plates. 

The articulating process could not be satisfactorily studied ; it appears, however, 
to be very robust for a shell of this size — the largest specimen observed being only 
12 mm. long. 

The only Brachiopod to which this species bears some structural resemblance is the 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 165 

Devonian Caraarospira eucharis (Hall).* The features common to both are the 
presence of a spondylium in the pedicle-valve, and that of a septum in the brachial 
one ; but here the spondylium extends anteriorly to about three-fourths of the length 
of the shell, whilst in Camarospira eucharis it is only about one-fourth of that length, 
and its supporting septum is much shorter. 

It must be understood that the reference of this species to the genus Camarospira 
is for the sake of convenience only, and cannot be substantiated without confirmatory 
evidence from other important characters. In any case, enough is known to show that 
it has no affinities to any of the smooth Camarophorids. 



StROPHOMENIDjE. 

Genus Schuchertella Girty. 

Schuchertella crenistria (Phillips). 
Davidson, loe. cit, pi. xxvi. 

A few fragmentary specimens of this species are in the collection. Dr G. H. Girty 
has fully explained his reasons for restricting the term Orthotetes to its original impli- 
cation, viz. to septate forms allied to Derby a Waagen ; and for the group of Strepto- 
rhynchus lens White, and of Spin/era crenistria Phillips, thus left anonymous, he pro- 
poses the generic term Schuchertella. 

In the present determination it is assumed that Phillips' type-specimen belongs to 
the aseptate form as interpreted by Davidson, t 

Orthotetid indet. (PI. II. figs. 23-23a.) 

A small Orthotetid, represented by a pedicle-valve and two brachial valves, is char- 
acterised by its biconvex shape, with the hinge-line shorter than the greatest width of 
the shell. The cardinal area of the pedicle-valve is very elevated, and the beak is dis- 
torted; the brachial valve has no area. The ornamentation consists of fine raised lines 
separated by broad sulci. 

This combination of characters gives it a superficial resemblance to Meekella (Ortho- 
tetina) olivieri (de Verneuil), but sections made in both valves show the total absence of 
the dental and septal plates characteristic of Meekella, and further, there is no median 
septum as in Derbya and in Orthotetes [emend. Girty]. Access to the details of the 
articulating process being impossible, I am unable to say whether this form should be 
considered as an aberrant Schuchertella or be referred to the genus Streptorhynchus. It 
would be interesting to prove the occurrence of a Streptorhynchus in the Cape Cherney 
Limestone, since the genus is said by Schellwien to have a far greater vertical range 

* Hall and Clarke, Palaeontology of New York, vol. viii., pt. ii., 1893, pi. 1. figs. 46-52, p. 82. 

t G. H. Girty, "The Guadalupian Fauna," U.S.G.S. Professional Paper No. 58, 1908, pp. 156-199. 



166 DR G. W. LEE ON 

than is usually admitted, having, according to him, made its first appearance in 
Devonian times.* 

Orthid^e. 

Genus Rhipidomella Oehlert. 

Rhipidomella michelini (L'Eveille). (PL I. fig. 16.) 
Davidson, loc. cit., pi. xxx. figs. 6-8. 

So far as external characters may be trusted, the numerous specimens which I refer 
to Rhipidomella michelini are undistinguishable from the west of Europe form. 

Genus Schizophoria King. 

Schizophoria aff. resupinata (Martin). (PI. I. fig. 17-1 7 a.) 

Compare Davidson, loc. cit., pi. xxx. fig. 1. 

A single pedicle-valve, globose and very transverse, with a shallow mesial sulcus, 
has the general appearance of Schizophoria resupinata, from which it differs in the 
nature of its ornamentation. The ribs are stronger and separated by wider sulci than 
in the British form ; near the margin the sulci are as broad as the ribs separating 
them. The shell is very thick. The pores and the pits corresponding to the spine-bases 
are disposed as in Schizophoria resupinata. 

Schizophoria sp. (PI. II. figs. 24-24c.) 

This species is represented by a complete specimen measuring : — 

Length . . . . .11 mm. 

Width . . . . . 13 „ 

Depth . . . . . 5 ,, 

and by two isolated valves. The ribbing is very fine, and the valves are non-sinuate 
and much depressed ; the cardinal area is short and low. The shell is very thin. 

A puzzling feature exhibited by the brachial valve after decortication is the hexa- 
plicate appearance of the muscular impression ; it may be, as suggested by my 
colleague Dr Ivor Thomas, who kindly looked at the specimen, that the posterior con- 
centric division is but the termination of unusual accretion of material below or about 
the cardinal process. The cast of the brachial valve of Orthis pulvinata Salter, shows 
■ in identical feature, due, as explained by Salter, to the impression of the " hinge-teeth " 
and of the cardinal process, the three happening to be of the same length in this species ; 

* E. Schellwien, loc. cit., p. 17 ; and " Beitrage zur Systematik der Strophomeniden des Oberen Palaeozoicum," 
Neues Jahrbuch, Bd. i., 1900, p. 5. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 167 

but in the present case, what is left of the shell in the cardinal region is more in 
favour of Dr Thomas' explanation.* 

The depressed shape of this shell gives it a greater resemblance to the form referred 
by Semenow and Moller to " Orthis striatula " than to any other Schizophoria, but 
in the Nowaja Semlja shell the hinge-line is shorter, t 

Production. 

Genus Chonetes Fischer de Waldheim. 

Chonetes papilionacea (Phillips). (PL IJ. fig. 25.) 
Davidson, loc. cit., pi. xxxv. figs. 3-5. 

The Cape Cherney form is undistinguishable from the flatter variety of the British 
Chonetes papilionacea. 

Dr T. F. Sibly has recently remarked that the term papilionacea embraces more 
than one form, and for the flat kind — as is the present one — he proposed the new name 
" compressa," a name very expressive of the general appearance of the shell, but 
unfortunately already applied to another species of Chonetes.% 

Chonetes (sp. plur. ?) (PL II., figs. 26-28a.) 

Numerous specimens of all sizes up to 12 mm. belong to one or more species, the 
general characters of which are as follows : — The shape is more or less semicircular, and 
the pedicle-valve evenly convex, but rather shallow ; the brachial valve is very concave, 
so that there is little space between it and the brachial one. The ribs increase by 
forking, but do not number more than about forty, on the larger specimens. Owing to 
the fragmentary condition of the material an accurate determination cannot be given, 
but it may be indicated here that where the shell's width does not exceed its length, 
and the shell is ornamented by few ribs, it bears a certain resemblance to the Devonian 
Chonetes armata (Bouchard), as figured by de Koninck,§ whilst the more transverse 
ones, with more numerous ribs, appear identical with the form provisionally described 
as Chonetes cf. crassistria by Dr A. Vaughan.|| 

A third type is represented by a specimen more globose than the above, with a com- 
paratively low area, as obtains, for instance, in Chonetes minuta (Goldfuss), from which 
it differs in the nature of its ribbing. IT 

* J. W. Salter, " Note on the Fossils from the Budleigh Salterton Pebble-Bed," Quart. Jour. Geol. Soc, 1864, 
p. 295, pi. xvii. fig. 8 ; Davidson, Brachiopoda of the Budleigh Salterton Pebble-Bed, Pal. Soc. 1881, pi. xli. fig. 11. 

t " Uberdie oberen devonischen Schichten des Mittleren Russlands," von P. Semenow und V. von Moller, Bull. 
Acad. Imp. Sciences, St Pe'tersbourg, 1863, p. 691, pi. ii. fig. 10. 

| Waauen, Pal. Indica : Salt Range Fossils, 1884, iv. p. 630 ; T. F. Sibly, " On the Faunal Succession in the 
Carboniferous Limestone of the Midland Area," Quart. Jour. Geol. Soc, 1908, p. 78, pi. i. fig. 7. 

§ Monographic des Genres Productus et Chonetes, 1847, pi. xx. fig. 14. 

|| " Palaeontological Sequence in the Bristol Area," Quart. Jour. Geol. Soc, 1905, p. 294, pi. xxvi. fig. 2. 

IT As interpreted by de Koninck, loc. cit., fig. 18. 
TRANS. ROY. SOC. EDIN, VOL. XLVII. PART I. (NO. 7). 25 



168 DR G. W. LEE ON 

Genus Productus J. Sowerby. 

Productus giganteus (Martin). 

The materials are fragmentary, but enough is at hand to show that more than one 
form may be represented. 

(I) A specimen which when complete must have measured about 15 cm. in width 
has the type of ribbing and the furrowed surface characteristic of the British examples 
figured by Davidson, pi. xxxviii. and xxxix. (loc. cit.). (2) A specimen about 12 cm. wide 
and very transverse, the surface of which is not furrowed, has the ribs separated by wider 
sulci than in the above. (3) A very large specimen, about 15 cm. wide, is less transverse 
than the two others, and has finer and more close-set ribs ; its surface is not furrowed. 
(4) Three small specimens have the Edelburgensis type of ornamentation. 

These four forms are too poorly represented to enable one to say whether they 
belong to distinct varieties (or species ?) or are but the different growth-stages of one 
and the same species. 

Productus antiquatus J. Sowerby. 
Mineral Conchology, pi. 317, figs. 5-6. 

The collection contains a large specimen, 7 cm. in width, agreeing very well with 
J. Sowerby's figure. 

This form and the following I cite separately, but without expressing an opinion as 
to their relationship to the Productus semireticulatus of authors. Speaking of the 
latter, Professor Tschernyschew has made the interesting remark that the flatness of 
its dorsal valve is a good criterion for distinguishing it from certain species of the group 
of Productus boliviensis d'Orbigny, to which it often bears some resemblance. 

Productus concinnus J. Sowerby. (PI. II. fig. 33.) 
Mineral Conchology, pi. 318, fig. 1. 

A strongly geniculated brachial valve seems referable to this species, the main 
characters of which are well exhibited, with this difference, that in the frontal portion 
the ribs are separated by wider sulci than appears to be the case in the British form ; 
but this may be an age-character, this being a large specimen 30 mm. in width. 

Productus longispinus J. Sowerby. (PI. II. fig. 31.) 

Davidson, loc. cit, pi. xxxv. figs. 5 and 12. 

The numerous specimens at hand belong to the slightly transverse, evenly convex 
type as figured by Davidson, fig. 5 (loc. cit.). One specimen only shows a slight con- 
vergence towards the lobate form. I have not been able to ascertain whether they 
possess the " niarginifera" ridge usually present in this species. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 169 

Productus longispinus? (var. ?). (PL II. fig. 32.) 

A few specimens are characterised by their fine ribbing and very convex visceral 
region. They resemble Productus spinosus J. Sowerby (Davidson, loc. cit., fig. 17), but 
the materials are too fragmentary to enable one to make a decided statement as regards 
their true affinities. 

Productus sp. indet. 

A fragmentary pedicle-valve, too incomplete for a detailed description, but belonging 
probably to the semireticulati, is suggestive, in the general style of its sculpture, of the 
form referred by Semenow and Moller to Productus carbonarius de Koninck.* It 
may, however, well be that this is but an appearance due to the imperfect state of 
preservation of the specimen. 

Productus cf. margaritaceus Phillips. (PI. II. fig. 29.) 

A pedicle-valve 10 mm. wide, and about 8 mm. long, is too small to permit of close 
comparison with the figures of Productus margaritaceus given by Phillips and 
Davidson, but on the other hand appears almost identical with the small form referred 
to this species by Dr A. ToRNQUiST.t 

The Upper Carboniferous Productus parvulus Nikitin [non Winchell, 1863] J has 
the same outline and ribbing, but differs from it in being more convex, with a broader 
visceral region on which are spines, here, however, very scarce and restricted to the 
cardinal slopes. 

Productus coatsi sp. nov. (PL II. figs. 35-35a.) 

This shell is longitudinally oval, and the hinge-line is a little shorter than the width 
of the shell. The pedicle-valve is gibbous, without a sinus ; the ears are very small. 
The beak is very prominent, and its extremity overlies the hinge-line. The brachial 
valve is just perceptibly concave, so that there is a large space between it and the 
pedicle-valve. 

The ornamentation of the pedicle- valve consists of tubercles which are very elongate 
and rib-like on the mesial and frontal portions, and rounded on the ears and cardinal 
slopes. The ornamentation of the brachial valve is very different, consisting of 
rounded pits. 

The specimen figured is a slightly distorted pedicle-valve, measuring : — 

Length . . . . .12 mm. 

Width . . . . . 9 „ 

Depth . . . . . 5 ,, 

* Loc. cit., Bull. Acad. Imp. Sciences St Pe'tersbourg, 1863, p. 704, pi. iv. fig. B. 

t Das fossilfiihrende Untercarbon am ostlichen Rossbergmassiv in den Vogesen, 1895, pi. xiv. fig. 4. 

I Nikitin, Mem. Com. Ge'ol. Russie, vol. v., No. 5, 1890, pi. i. figs. 13-14. 



170 DR G. W. LEE ON 

Another specimen, 1 1 mm. long, is partly broken, but shows both valves in apposition ; 
the maximum distance between the two valves is 5 mm. 

The general appearance of the shell is very much like that of Prodnctus youngianus 
Davidson, a shell which differs from the present one in having a very concave brachial 
valve closely following the curves of the pedicle-valve ; besides, the brachial valve of 
Productus youngianus is ornamented by ribs, the place of which is taken here by pits. 
The nature of the ornamentation on the ears and cardinal slopes can also serve to dis- 
tinguish it from Productus youngianus. 

The Devonian Productus dissimilis de Koninck,* is remarkably like Productus 
brucei as regards the ornamentation of the pedicle-valve, but it is a less elongate shell, 
with a differently ornamented brachial valve. 

Productus sp. indet. 

A specifically indeterminable fragment exhibits the ribbing typical of Productus 
scabriculus (Martin). 

Productus cherneyensis sp. nov. (PI. II. figs. 37-376.) 

No specimens showing both valves in apposition are at hand, but the pedicle- valve 
described here is sufficiently characterised to merit a distinct appellation. It is semi- 
circular, measuring 17 mm. in length and 19 mm. in width. The visceral portion is high 
and narrow ; the ears are well defined and the beak is much incurved, overlying the 
hinge-line. The surface is ornamented with ten broad concentric folds bearing thick, 
rounded tubercles separated by wide intervals. Between the folds, numerous concentric 
growth-lines are visible to the naked eye. The shell-substance is very thin. 

Two small specimens, 5 mm. long, show the same characters, the thick tubercles 
being very pronounced for such small shells. 

This species probably belongs to the group of Productus fimbriatus J. Sowerby, 
but it differs from most of its members in having a narrower visceral cavity, and fewer 
tubercles, which are moreover rounded, not elongate as in Productus fimbriatus. These 
characters are well shown in a species which Dr A. Vaughan provisionally describes as 
Productus cf. fimbriatus,^ with the remark that it probably represents a link between 
the fimbriate and the aculeate Producti. Productus cf. fimbriatus may possibly be 
identical with the present one, but the specimen figured by Dr Vaughan is more elongate 
than the Cape Cherney form. 

Productus elegans M'Coy. (PI. IT. figs. 38-386.) 

M'Coy, British Palaeozoic Fossils, 1855, p. 460, pi. iii.H, fig. 4. 
This is the most abundantly represented Productus in the collection, and it is 
undistinguishable from its British representatives. One specimen, showing part of both 

* Monor/raphie des Genres Productus et Chonetes, p. 147, pi. xvi. fig. 5. 
t Quart. Jour. Geol. Soc, 1906, p. 308, pi. xxx. fig. 6. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 171 

valves in apposition, reaches a size exceeding that commonly attained by the species ; 
when complete it probably" measured some 40 mm. in length. As this large specimen 
shows no trace of the sinus characteristic of Productus punctatus (Martin), we have here 
another conclusive proof of the fact that Productus elegans is not to be considered as 
the young stage of Martin's species. 



Productus keyserlingianus de Koninok, var. (PL II. figs. 36-36a.) 
De Koninck, Monographie du Genre Productus, 1849, p. 134, pi. xiv. fig. 6. 

Four specimens, two of which have both valves in apposition, differ from Productus 
keyserlingianus as described by de Koninck in having fewer tubercles ; they are other- 
wise quite like the Belgian form. 

The specimen figured under this name by M. Krotow is remarkable for an unusually 
large number of tubercles.* 

Productus spinulosus J. Sowerby. (PL II. fig. 30.) 

Davidson, loc. cit., pi. xxxiv. figs. 18, 19. 

The materials are fragmentary, but abundant ; they are quite characteristic, with the 
exception of one or two specimens which appear to be slightly flatter than the type as 
refigured by Davidson. 

Sub-genus Proboscidella Oehlert. 

Productus (Proboscidella) nysti de Koninck. (PL II. figs. 34, 34a.) 
De Koninok, loc. cit., p. 65, pi. xiv. fig. 5. 

Materials indifferently preserved and belonging to this species might easily be taken 
for brachial valves of other Productids, but the specimen in the collection is in a fair 
state of preservation, consisting of a nearly complete pedicle- valve with part of the other 
valve still in apposition. 

Its proportions and ornamentation, and its short bent margin, are exactly as depicted 
by de Koninck. 

Incert^e Sedis. 

The collection contains two fragmentary pedicle-valves of a Productoid shell, 
characterised by its greatly depressed shape and the extreme thinness of the test. The 
ornamentation consists of broad longitudinal folds bearing a few obtuse tubercles, with 
indications of wrinkling on the ears. 

* Mem. Com. Ge'ol. Russie, vol. vi., 1888, pi. i. fig. 20. 



172 1>R G. W. LEE ON 

Class LAMELLIBRANCHIATA. 
Genus Pal^eolima Hind. 

Palseolima aff. simplex (Phillips). (PI. II. fig. 45.) 

Hind, Monograph of the British Carboniferous Lamellibranchiata, Pal. Soc, 1903, vol. ii. p. 39, 

pi, xix. tigs. 24-27. 

A right valve measuring 6 mm. dorso-ventrally and about 6 mm. antero-posteriorly 
is very much like the British form as interpreted by Dr Hind, but its ribs are not 
appreciably flattened on the anterior side, and in this one respect differs from that form. 

From the Upper Carboniferous Palasolima retifera (Shumard) it differs in having 
fewer ribs. 

Genus Aviculopecten M'Coy. 

Aviculopecten clathratus (M'Coy). 

Hind, loc. cit., p. 82, pi. xv. figs. 1-7. 

A left valve measuring approximately 8 mm. dorso-ventrally agrees in every respect 
with Dr Hind's re-description of the species. 

Aviculopecten planoclathratus (M'Coy). 

Hind, loc. cit., p. 91, pi. xv. figs. 8-12. 

The shape and ornamentation of this specimen, which measures 10 mm. dorso-ventrally 
and 10 mm. antero-posteriorly, are exactly as in the British form, only the radial 
ornamentation is more accentuated ; but this may be accounted for by the excellent 
state of preservation of the surface characters in the Nowaja Semlja shell. 

Aviculopecten ? sp. indet. 

A badly preserved pectiniform shell, measuring 30 mm. approximately in its two 
dimensions, probably belongs to this genus, but not to either of the above species. 

Genus Edmondia de Koninck. 

Cf. Edmondia m'coyi Hind. 
Hind, loc. cit., vol. i., 1899, p. 329, pi. xxxvi. figs. 23-29. 

A left valve, somewhat obscure and incomplete, belongs probably to a species of the 
group of Edmondia scalaris (M'Coy), and its affinities to Edmondia m'coyi Hind, are 
indicated by the shape of the broad concentric ridges, which point to a transversely 
subquadrate shell rather than to an obliquely suboval one. It measures approximately 
15 mm. dorso-ventrally and 20 mm. antero-posteriorly. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 173 

Genus Conocardium Bronn. 

Conocardium rostratum (Martin). (PL II. fig. 40.) 

Hind, loc. cit., vol. i., 1900, p. 453, pi. li. figs. 6-9. 

This species is represented by two small specimens quite complete but for part of 
the anterior portion. The posterior rostrum of the specimen figured is in the collection, 
but could not conveniently be photographed along with it. This specimen measures 
3 "5 mm. dorso-ventrally and 3 mm. from side to side, and owing to its small size I am 
doubtful as to the degree of importance to be attached to the fact that the lower border 
is not concave, as in the form figured by Dr Hind, loc. cit., fig. 7, but is rather slightly 
convex, as in fig. 6, ibid. ; this may be only an appearance due to the incompleteness of 
the anterior end. 

Conocardium alseforme (J. de C. Sowerby). (PI. II. figs. 41, 41a.) 
Hind, loc. cit., vol. i., 1900, p. 460, pi. liv. figs. 1-10. 

Conocardium, alseforme is represented by seven specimens in all stages of growth 
up to about 10 mm. dorso-ventrally, and apparently undistinguishable from the British 
form, with this exception, that the larger one is slightly less inflated than Sowerby's 
type. At all events, it seems that the degree of gibbosity of the shell varies even in 
typical Lower Carboniferous districts — since a specimen of Conocardium alseforme 
figured by de Koninck is appreciably compressed laterally.* 

Class SCAPHOPODA. 
The collection contains a specifically indeterminable Dentaliid. 

Class GASTEROPODA. 

The Gasteropod fauna of the Cape Cherney limestone is very varied, but the greater 
number of the species is represented by extremely small specimens, as if some 
pauperising influences had arrested their normal development. In those cases where 
the types of the species which I believe to be present here are actually minute forms, 
my identifications will probably be found to be fairly correct, but it is not without some 
misgiving that I have made bold to compare some of these small fossils with common 
west of Europe species, although these have been founded on large specimens. Yet 
resemblance is often so striking — except in matter of size — that the latter course is 
perhaps after all the safest. 

* Faune du Calcaire Carbonifere de la Belgique, 1885, vol. v. pi. xviii. fig. 17. 



174 DR G. W. LEE ON 

Pleurotomarid^e. 

Genus Mourlonia de Koninck. 

Mourlonia Isms (M'Coy). (PI. II. figs. 44-446.) 

M'Coy, Synopsis of the Carboniferous Limestone Fossils of Ireland, 1844, p. 41, pi. v. fig. 15. 

This very depressed species appears to be represented by two small specimens 
measuring 6 mm. in diameter. The slit-band is perhaps slightly wider than is the case 
in the type-figure. 

Genus Murchisonia d'Archiac and de Verneuil. 

Murchisonia aff. archiaciana de Koninck. (PI. II. fig. 42.) 
De Koninck, Faune du Galcaire Carbonifere de la Belgique, vol. iv., 1883, p. 16, pi. xxxiv. figs. 5-6, 19-20. 

De Koninck's species was founded on a much larger form. The fragmentary 
specimen which I compare to it is much smaller, the length of the four whorls com- 
posing it being only 6*5 mm., and the diameter of the last whorl 3 mm. The tapering 
is the same as in the Belgian species, and the ornamentation is similar, but there are 
four keels on each side of the sinual band, whilst in Murchisonia archiaciana four 
keels are stated to be above the band, and six or seven below it ; the latter are also less 
pronounced. 

Bellerophontid^e. 

Genus Bellerophon de Montfort. 

Bellerophon hiulcus (Martin). 

J. db C. Sowerby, Mineral (Jonchology, 1825, pi. 470, fig. 1. 

Bellerophon hiulcus is represented by a few specimens, the largest of which does not 
exceed 12 mm. in diameter. 

Bellerophon aff. tenuifascia J. de C. Sowerby. 

J. de C. Sowekby, Mineral Conchotomy, 1825, pi. 470, figs. 2-3. 

The ornamentation and the fine, thread-like ridge are as in the British species, from 
which the specimen at hand — measuring 18 mm. in diameter — differs in being less 
globose, the periphery being somewhat flattened, especially on the last part of the whorl. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 175 

EUOMPHALID^E. 

Genus Straparollus de Montfort. 
Straparollus dionysi de Montfort. 
De Koninck, loc. cit., vol. iii., 1881, p. 120. 

This species, as interpreted by de Koninck, is represented by a few specimens 
greatly dwarfed, the largest of which is only 1 mm. in diameter. 

Straparollus pileopsideus (Phillips). 

Phillips, Geology of Yorkshire, vol. ii., pi. xiii. fig. 6. 

The collection contains numerous specimens of a small species of Straparollus which 
in every detail except size seems identical with Phillips' species ; they average only 
5 mm. in diameter. 

Pyramidellid^e. 

Genus Loxonema Phillips. 

Loxonema Isevigatum (Eichwald). (PL II. fig. 46.) 

Lethxa Rossica, 1860, p. 1117, pi. xlii. fig. 6. 

The specimens which I refer to this species agree very well with Loxonema 
Isevigatum, in particular in the smooth surface, the shape of the whorls, the linear 
suture, and the tapering of the shell. They are, however, much larger, the specimen 
figured, for example, measuring 4 mm. across the last whorl, a dimension which far 
exceeds that of Eichwald's type. 

Loxonema suturale (Phillips) has the same general appearance, but its suture is 
wide and deeply excavated. If my identification proves correct, this will be the only 
Gasteropod in the collection actually larger than the type of the species to which it is 
referred. 

Loxonema sp. a. (PI. II. figs. 48, 48a.) 

An incomplete specimen of four whorls, the total length of which is 4 "5 mm., is 
characterised by a wide spiral angle and an ornamentation consisting, on the juvenile 
whorls, of numerous fine lines, equidistant and raised, which, on the last whorl, are 
arranged in sets of three separated by a smooth interval, the central line being a little 
stronger than the two bordering it. 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 7). 26 



176 DR G. W. LEE ON 

Loxonema cf. anglicum d'Orbigny. 
Phillips, Palaeozoic Fossils of Cornwall, 1841, fig. 188. 

Four specimens, very small and fragmentary, must be placed in the group of 
Loxonema rugiferum (Phillips), but the whorls are higher than in this species, thus 
pointing to closer affinities to Loxonema anglicum d'Orbigny. Another character 
they appear to have in common with it, is that the ribs are continuously uniform in 
size, showing no tendency to become evanescent in their upper part. This latter 
feature has already been pointed out by the Rev. G. F. Whidborne.* 

The name Loxonema anglicum was given by d'Orbigny t to a Devonian species 
which Phillips had described in 1841 (loc. cit.) as a Devonian representative of his 
Carboniferous Melania rugifera, from which it differs in the characters indicated. 

The materials are too fragmentary to determine whether they really belong to the 
Devonian species, which has, however, already been recorded from Carboniferous rocks, 
the name fio-uring in Struve's list of Carboniferous fossils from the Moscow basin. 



o 



Loxonema sp. b. (PL II. fig. 43.) 

The largest of three fragments is composed of four whorls measuring 3 '5 mm. in 
length. The ornamentation consists of straight, sharp equidistant costse, more numerous 
on the younger whorls than on the later ones, and belongs to the type exhibited in 
Loxonema strigillatum de Koninck (loc. cit., 1881, pi. vi. fig. 22) and in Loxonema 
wisherse Krotow.J But the Nowaja Semlja shell differs from these in having the whorls 
less convex and a more shallow suture. In shape it resembles Loxonema semicostatum 
Meek and Worthen ; § but in the latter the costse increase in number in the older part 
of the shell, whilst in this case it is the reverse. 

Loxonema? sp. c. (PL II. fig. 49.) 

A nearly complete specimen, measuring 5*5 mm. in length and If mm. across the 
body-whorl, has doubtful generic affinities, since it is bucciniform, with a large body- 
wborl t and an apparently twisted columella, as in certain species of Macrocheilina, but 
the ornamentation is like that of a Loxonema. Since the state of preservation is such 
as to preclude the possibility of determining whether the columella was callous or not, 
its true generic position must remain doubtful. 

The first two whorls are unornamented ; the third and the fourth have a few broad 
transverse folds, whilst the ornamentation of the body-whorl is totally different, consist- 
ing of straight raised lines, closely set and very numerous. 

* A Monograph of the Devonian Fauna of the South of England, Pal. Soc, 1896, vol. iii. p. 43. 
t I' rod nunc tic I'tili'ttiitoloyie, vol. i., 1850, p. 62. 
I Mem. Com. G4ol. Russie, vol. vi., 1888, pi. i. fig. i. 
§ Geol. Survey Illinois, vol. v., pi. xxix. fig. 2. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 177 

Its ornamentation distinguishes it from Loxonema ? buccinoideum de Koninck (loc. 
cit., 1881, pi. vi. figs. 12-13), a species of doubtful generic affinities but of a similar 
shape, as is also Loxonema semicostatum de Koninck \non Meek and Worthen]. 

LlTTORINIL\E. 

Genus Turbonitella de Koninck. 

Turbonitella biserialis (Phillips). (PL II. figs. 47, 47a.) 

Phillips, Geology of Yorkshire, pi. xiii. figs. 10, 11. 

This species is represented by four small specimens, the largest of which is only 5*5 mm. 
in length. The small ones have the ornamentation of Phillips' Turbo semisulcatus 
{loc. cit,, fig. 10), and the greater portion of the last whorl of the largest one has the 
two sets of alternating ribs as typically developed in Turbo biserialis (loc. cit., fig. 11), 
its beginning showing the transition of the " semisulcatus " type to the " biserialis " 
type. This is also exhibited on the end of the last whorl of the intermediate specimen. 

Genus Portlockia de Koninck. 

Portlockia lacordairei ? (de Koninck). 
De Koninck, loc. cit., 1883, pi. xxv. figs. 17-19. 

Portlockia parallela ? (Phillips). 
Phillips, Geology of Yorkshire, pi. xvi. fig. 8. 
Each of these two species seems to be represented by a fairly complete but ex- 
tremely minute specimen, in all respects like the west of Europe form, except in size. 

Capulhle. 
Genus Orthonychia Hall. 

Orthonychia sp. (PL II. figs. 39, 39a.) 

Without expressing an opinion as to the value of the generic term Orthonychia, I 
place under it a very small (6 mm. from apex to frontal margin) capuliform shell which 
exhibits the features said to be characteristic of 'the genus, that is to say, the apex is 
incurved to a slight extent only, and shows no tendency to be spirally enrolled. Its 
specific characters are as follows : — The slopes are steep and meet under an acute angle 
forming a ridge bordered on each side by a very faint carina ; a shallow groove extends 
from the apex to the margin, and divides the left side into two equal portions ; this 
groove is apparently absent on the right side ; the surface is covered with undulating 
lines of growth. 



178 DR G. W. LEE ON 

Comparison with other species is difficult, since it is very small and there is 
nothino- to indicate the stage of growth of the specimen, but in many respects it bears 
a general resemblance to Orthonychia uncum (Meek and Worthen), # a species which 
American palaeontologists consider to be a synonym of Orthonychia acutirostris 
(Hall).t 

Class CRUSTACEA. 

Trilobita. 

Genus Phillipsia Portlock. 

Phillipsia eichwaldi ? (Fischer de Waldheim). 
Woodward, Monograph of the British Carboniferous Trilobita, 1883, p. 22, pi. iv. 

A pygidium and a fragment of the head, seen from the under side, are possibly 
referable to this species. The under side of head is very much like that figured by 
Dr H. Woodward (loc. cit., fig. 8). 

I am indebted to Dr B. N. Peach, F.R.S., for helping me in this determination. 

The following are generically indeterminable, on account of their imperfect 
preservation : — 

Sp. a. 

A fragment of pygidium characterised by the extreme convexity of the axis and 
the corrugated appearance of the somites, as obtains, for instance, in Griffithides 
globiceps (Phillips). 

Sp. b. (PL II. figs. 50, 50a.) 

A fragmentary pygidium, to which are attached three axial thoracic somites. 
Only the axis, consisting of fourteen somites, and a few of the pleural segments are 
preserved. 

The three anterior axial rings of the pygidium are ornamented, on their median 
portion, by a single row of strong tubercles ; laterally, these tubercles are smaller and 
disposed in a double row. On the other rings the tubercles are in one row on the whole 
surface, and on every third ring the central tubercle is much more prominent than those 
next to it. The pleural segments are less numerous but broader than the axial rings, 
and are separated by wide and deep sulci. Each segment is longitudinally divided by 
a well-defined furrow in two sub-equal ridges, of which the anterior one is the broader ; 
both are ornamented by a row of closely set tubercles. 

The fragments of thoracic segments are quite smooth, and are well defined. 

The specimen is too poorly preserved to permit of any suggestion as to its 

* Geol. Survey Illinois, vol. v. 1873, pi xvi. fig. 1. 
t Geol. Survey Iowa, 1858, pi. xxiii. fig. 14. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 179 

relationships, but it might be mentioned that it appears to belong to an undescribed 
species bearing a certain resemblance to Brachymetopus ouralicus (de Verneuil), as 
figured in Murchison's Russia, from which it differs, however, in the proportions 
obtaining between the divided parts of the lateral segments, and in the nature of the 
ornamentation on the three anterior axial rings. 

Ostracoda. 

Thin sections of the limestone show that carapaces of Ostracods are common, but 
only a few could successfully be freed from the matrix. These are in excellent state of 
preservation, the finer details being well exhibited. They appear to be undistinguish- 
able from the following species : — 

Leperditia okeni (Miinster). 
Bairdia curta (M'Coy). 
Cytherella ? inflata (Miinster). 

General Remarks. 

Considering that the fossils described in this paper were found in a few fragments of 
limestone, and these from a single bed, the great variety of groups and species is really 
surprising ; representatives of most of the Lower Carboniferous invertebrate groups are 
met with, with the exception of those which are usually absent from limestone strata — 
the Cephalopods, for instance. 

So far as the evidence at hand goes, it seems probable that the bed yielding these 
fossils was deposited in comparatively clear water at some distance from land, but not 
necessarily at any great depth. The Protozoa, the Corals, the Crinoids, and the 
Brachiopods grew in the midst of conditions eminently suitable for their development ; 
it is therefore all the more remarkable that the Lamellibranchs and the Gasteropods 
should be represented by so many dwarfed individuals. At the same time, these 
dwarfing influences did not prevent the multiplication of species in the groups affected. 

As regards the stratigraphical position of the Cape Cherney limestone, the assem- 
blage as a whole points to a Lower Carboniferous age, whilst the presence of many 
typical species warrants us in assuming that the bed is probably in a series homotaxial 
with the Productus giganteus zone of Russia or Upper Visean of Western Europe. 

The collection does not contain any of the species characterising the upper divisions 
of the Carboniferous system, but at the same time not a few of the species cited in this 
paper have been recorded from horizons higher than the Visean. However, with our 
extended knowledge of the Upper Palaeozoic formations, competent palaeontologists have 
found it necessary to extend greatly the range attributed to some of the more common 
fossils ; these forms have probably undergone certain evolutionary changes, but since 
these are often not discernible, new names have not been given to the apparently 



180 BR G. W. LEE ON 

unmodified descendants of the forms originally described from lower horizons. Through 
the kindness of Professor Th. Tschernyschew, I am enabled to mention a striking 
instance of persistence in the most familiar of Lower Carboniferous fossils : he found a 
specimen of Productus giganteus in the Spirifer mosquensis limestone of the Timan, a 
region where the Lower Carboniferous is absent and the Mosquensis zone transgresses 
on the Devonian. Vice versa, Upper Carboniferous species may have appeared earlier 
than usually supposed, and, to mention an example taken from this country, Dr 
Vaughan cites Derby a grandis Waagen, from the top of the Carboniferous limestone 
at Loughshinny, Ireland.* 

Bearing these facts in mind, one might hesitate to assign a definite age to a limited 
collection of Carboniferous fossils ; but in the present case the evidence is very strongly 
in favour of a correlation with the Giganteus zone, since the number of species typical 
of this zone is as large as that from any bed in a typical Visean locality, not to mention 
the complete absence of typical Upper Carboniferous forms. 

In the following table I give a list of the species found in the Cape Cherney lime- 
stone, with an indication of their occurrence in the Lower Carboniferous beds of Russia 
and Western Europe. Their occurrence in higher horizons is also indicated, but these 
data have been compiled from various sources, and needless to say no Upper Carboni- 
ferous locality is known where so many Lower Carboniferous forms would occur together. 
It must also be understood that many of the species quoted from Upper Carboniferous 
beds are probably not strictly identical with the Lower Carboniferous forms the names 
of which they bear, for the reason stated above. Had they all been described instead of 
being simply cited in lists, I should perhaps have been justified in dispensing with many 
of these comparisons. 

As regards the relation of the fauna to its distribution in space, it seems to be very 
closely allied to that of the Giganteus zone of the Urals and Central Russia, which is 
practically the same as that of Western Europe. The list of fossils described in this 
paper has many points in common with those given respectively by Messrs Tscherny- 
schew, Krotow, Krasnopolsky, etc., in various monographs treating of the Carboni- 
ferous of the Urals. t 

It is true that, of the districts surveyed by these authors, the most septentrional one, 
viz. the district of Tscherdyn and Ssolikamsk described by M. Krotow, lies some ten 
degrees south of Cape Cherney ; but Professor Tschernyschew kindly informs me that 
I ' roductus giganteus has been found much farther north, in the basin of the river Adzva, 
a branch of the Oussa. Thus, although marine Lower Carboniferous beds' have not yet 
been proved to exist in the extreme north of the Ural range (Pai-Khoi), it appears 
probable that the Cape Cherney limestone was deposited in a sea situated in the con- 
tinuation of the Uralian geosyncline. At all events, the Lower Carboniferous sea did 
not apparently extend farther west — over the Timan — and the Cape Cherney fauna 

* Quart. Jour. Geol. Soc, 1908, p. 446. 

t M6n. Com. Geol. Russie, t. iii., 1889, t. vi., 1888, and t. xi., 1889. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 181 

shows no evidence of having inhabited an isolated basin, nor does it bear any affinities 
to the Lower Carboniferous fauna of North America. 

A still closer agreement obtains between it and the Lower Carboniferous fauna of 
the Moscow and Donetz basins * and of Western Europe, but this is obviously due to 
these regions being more accessible than the Northern Urals to exhaustive palseonto- 
logical studies. Comparison with Struve's list shows a striking resemblance to the 
lower portion of the Giganteus zone in the Moscow basin (" Kalkstein mit Stigmarien "), 
but of course this may be due to nothing more than mere similarity in the conditions of 
deposition. 

In spite of the fact that Cape Cherney lies very far from the typical Lower Carboni- 
ferous districts, its fauna shows very few characters peculiar to itself. The maximum 
isolation appears to obtain in the Corals, but it is still possible that those forms which 
are new may at some future date be found in the Northern Urals.t 

The Brachiopods are closely allied to those of the typical Lower Carboniferous 
districts, and if I have ventured to establish a few new species and left others unnamed, 
the greater number of them are undistinguishable from their representatives as developed 
elsewhere. The other groups are likewise represented by forms which appear to be 
remarkably cosmopolitan, and call for no special remarks. 

In the following table the sign + denotes the presence of species believed to be 
identical with those described in this paper, and cf. indicates that nearly allied or 
vicarious species are probably represented. 

* See Struve's list in " Schichtenfolge in den Carbonablagerungen im stidlichen Theil des moskauer Kohlen- 
beckens," Me'm. Acad. Imp. St Petersbourg, 1886 ; and " Excursion XVI.," by Messrs Tschernyschew and Loutougin, 
in Guide des Excursions du VII me Congres Ge'ologique International, 1897. 

t See notes by Mr R. G. Carruthers. 



[Table. 



182 



DR G. W. LEE ON 



Table showing the Fauna of the Cape Cherney Limestone and its Horizontal 
and Vertical Distribution in Russia and Western Europe. 







^ 


to 












on Beds of 

Murajewna 
Carboniferous ' 
al Kussia). 


.3 

i-t 03 


CO 

p 


CO 

3 

P CD 


co j; cs 




CD 

a 

CD 


§§ 

-SO 


CD 
O co" 

^ f- 1 


irbonifei 
sean). 
n Europ 


n c3 « 




CD 

Pa 

o 


Transiti 

Malewka- 

(" Devono-( 

of Centi 


&1 

u a 

5 o 
o 


cS f— , 

CD 

O 


Lower Cf 

(Vi 
Wester 


Higher 
Arctic a 
Europe. 


Endothyra globula (Eich.) 


+ 




+ 


+ 


cf. 




Spirillina 1 sp. 


+ 












" Crihrostomum " gracile Moll. 


+ 




+ 






+ 


Tetrataxis gibba Moll. 


+ 




+ 








Arclisediscus karreri Brady . 


+ 




+ 


+ 


+ 




Fusulinella struvei Moll. 


+ 




+ 


+ 






? Lagena sp. .... 


+ 












Genosplixra sp. a 


+ 








cf. 




? sp. b . 


+ 








cf. 




Carpospheeral sp. .... 


+ 








cf. 




1 Cenellipsis sp. 


+ 








cf. 




Renieria clavata Hinde 


1 








+ 


+ 


Aulophylhim sp. 


+ 












Lophophyllum cherney ense sp. nov. . 


+ 












sp. .... 


+ 




• 








Campophyllum carinatum sp. nov. . 


+ 












,, kungurense (Stuck.) . 


+ 










+ 


Carcinophyllum sp. . 


+ 












Hexaphyllia m'coyi (Duncan) 


+ 








+ 




Crinoid remains .... 


+ 












Stenopora sp. 


+ 








cf. 


cf. 


Polypora papillata M'Coy 


1 




+ 


+ 


+ 


+ 


*Dielasma lenticulare de Kon. 


+ 








+ 




„ gillingense (Dav.) . 


+ 






cf. 


+ 


cf. 


Spiriferina insculpta (Phill.) 


+ var. 




+ 




+ 




„ odoplicata (J. de C. Sow.) 


+ 


+ 


+ 


cf. 


+ 


+ 


„ sp. 


+ 












„ luminosa M'Coy 


1 








+ 


cf. 


Spirifer Imulcatus J. de C. Sow. 


% 




+ 


+ 


+ 


cf. 


„ triradialis Phill. 


1 




+ 




+ 




Martiniopsis 1 sp. 


+ 












Martinia sp. [cf. Sp. planata Phill.] 


+ 








cf. 




„ brucei sp. nov. 


+ 












,, sp. [aff. rhomboidalis M'Coy] 


+ 








cf. 




Reticularia imbricata (J. Sow.) 


1 








+ 




.Squamularia sp. a . 


+ 




cf. 


cf. 


cf. 


cf. 


sp. b . 


+ 




cf. 


cf. 


cf. 


cf. 


Ambocodia urei (Flcm.) 


1 




+ 




+ 


+ 


Rhynchonellids .... 


+ 




cf. 


cf. 


cf. 


cf. 


Athyrids ..... 


+ 




cf. 


cf. 


cf. 


cf. 


' 'amarospiral sp. . 


+ 












Schuchertella crenistria (Phill.) 


+ 




+ 


+ 


+ 




Orthotetid ..... 


+ 










cf. 


Bhipidomella michelini (L'EveilM) . 


+ 




+ 


+ 


+ 


+ 


Schizophoria resupinata (Mart.) 


var. 




+ 


+ 


+ 


+ 



* Tournaisian species. 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 



183 



Table showing the Fauna of the Cape Cherney Limestone — continued. 







s 













tk 


eds of 
ajewna 
niferous 
issia). 


> "m 

o l— 

rs cd 


t» 

3 
p 

c2 


liferous 
rope. 


zons. 

astern 
.issia. 




5 


f 3 SJtS 


a a 
o o 


'3 - 

o =" 


2 «m 


•g W « 




O 

CD 

eg 

Q 


Transition 

Malewka-M 

(" Devono-Cai 

of Central 


-go 

M in 

o 


"2 2 

CS h^ 
CD 

o 


Lower Carl 

(Visea 

Western 1 


Higher H 
Arctic and 
Europe. 


Sclrizophoria sp. . 


+ 


cf. 










Chonetes papilionacea (Phill.) 


+ 




+ 


+ 


+ 




sp. . 


-t- 








cf. 




Produdus giganteus (Mart.) . 


+ 




+ 


+ 


+ 


+ 


,, antiguatus J. Sow. 


+ 




cf. 


cf. 


+ 


cf. 


,, concinnus J. Sow. 


-j- 




+ 




+ 




,, longis-pinus J. Sow. 


+ 




+ 


+ 


+ 


+ 


?var.?. 


+ 








cf. 




,, margaritaceus Phill. [Tornquist] . 


+ 








+ 




,, coatsi sp. nov. 


+ 












„ scdbriculus (Mart.) 


1 1 




+ 




+ 


+ 


„ clierneyensis sp. nov. 


+ 








cf. 




,, elegans M'Coy 


+ 




+ 




+ 


+ 


,, keyserlingianus de Kon. . 


var. 








+ 


cf. 


,, spinulosus J. Sow. 


+ 








+ 


cf. 


,, sp. [ = Pr. carbonarius Sem. and 














Moll, non de Kon.] 


1 1 


+ 










„ nysti de Kon. 


+ 




+ 




+ 


+ 


„ sp. indet. .... 


+ 












Palxolima simplex (Phill.) . 


cf. 








+ 




Aviculopeden dathratus (M'Coy) 


+ 








+ 




„ planudalhratus (M'Coy) 


+ 








+ 




., sp. indet. 


+ 












Edmondia maceoyi Hind. 


? 








+ 




Conocardium rostratum (Mart.) 


+ 








+ 




„ alxforme (J. de 0. Sow.) 


+ 




+ 


+ 


+ 




Dentaliid ..... 


+ 












Mourlonia leevis (M'Coy) 


+ 




+ 




+ 




Murchisonia ardiiaciana de Kon. 


cf. 




+ 




+ 




Bellerophon hiulcus (Mart.) . 


+ 








+ 


+ 


,, tenuifascia J. de C. Sow. 


cf. 








+ 




Straparollus dionysi de Montfort 


+ 




+ 


+ 


+ 


+ 


,, pileopsideus (Phill.) 


+ 




+ 




+ 


+ 


Loxonema Ixvigatum (Eich.) 


+ 


+ 


+ 








sp. a 


+ 












„ anglicum d'Orb. . 


cf. 




+ 








sp. b . 


+ 












1 sp. c 


+ 








cf. 




Turbouitella biserialis (Phill.) 


+ 






+ 


+ 




Porilockia lacordaireana (de Kon.) . 


+ 








+ 




„ parallela (Phill.) . 


+ 








+ 




Orthonydiia, sp. . 


+ 












Phillipsia eichivaldi (Fischer) 


1 




+ 




+ 




Trilobites, indet. (a) . 


+ 












„ (6). 


+ 












Leperditia okeni (Minister) . 


+ 








+ 




Bairdia curta (M'Coy) 


+ 








+ 


+ 


Cytherella inflata (Miinster) . 


+ 








+ 





TRANS. ROY. SOC. EDIN, VOL. XLVII. PART I. (NO. 7). 



27 



184 DR G. W. LEE ON 

EXPLANATION OF PLATES. 
Plate I. 

Fig. 1. LopliophyUum cherneyense sp. nov. Transverse section of a mature example. Epitheca entire 
except on the right hand side. x 3/2. (The orientation is incorrect, the cardinal fossula being on the right- 
hand side of the figure, instead of the upper side.) 

Fig. 2. Median vertical section of another specimen. Showing the vesicular tabulae and 

dissepiments and the thickened columellar septum. x 3/2. 

Figs. la-Id. Serial transverse sections from the same specimen as fig. 2. Epitheca only 

complete in fig. 2ft. The LopliophyUum stage, with the septa extending to the columella, is seen in 
figs. 2c and 2d. x 3/2. 

Fig. 3. Campophyllum carinatum sp. nov. Median vertical section of a mature example. The carinas, 
running upwards and inwards at a steep angle, are visible over part of the dissepimental area. Epitheca 
worn away, x 2. 

Fig. 3a. Transverse section cut immediately below above. The major septa show a 

maximum extension towards the centre of the corallum. The carinae are best seen in the right-hand bottom 
corner. Epitheca worn away. x 2. 

Fig. 4. Transverse section of another example, showing a minimum extension of the 

major septa towards the centre, another section from this specimen agreeing with fig. 3ft. Epitheca almost 
entire. x 2. 

Fig. 5. ■ ■ Median vertical section of another examjDle, showing the highly arched tabulae. 

Carinas visible at the right-hand top corner. Epitheca practically entire. x 2. 

Fig. 6. A young example, showing the general external characters. x 1/1. 

Fig. 7. LopliophyUum sp. Transverse section. In the upper half the epitheca is entire, x 1/1. 

Fig. 7ft. Median vertical section from the same specimen as above, x 1/1. 

Fig. 8. Carcinophyllum sp. Transverse section just below floor of the calyx. Epitheca preserved in 
right-hand upper corner. x 1/1. 

Fig. 8ft. Median vertical section from same specimen. Epitheca just worn away, x 1/1. 

Fig. 86. ■ Transverse section from younger part of same specimen. Epitheca almost entire. 

xl/1. 

Fig. 9. Hexaphyllia m'coyi (Duncan). Median vertical section showing the thick wall (on right-hand 
side of figure) and the closely set tabulae. x 1/1. 

Fig. 10. Dielasma tenticulare de Koninck. Lateral view. xl/1. 

Fig. 10ft. Ventral view of same specimen. x 1/1. 

Fig. 106. Dorsal view of same specimen. x 2. 

Fig. 10c. ? Ventral view of a large specimen. x 1/1. 

Fig. 11. Spiriferina insculpta (Phillips), var. Dorsal view. The beak is broken. x 1/1. 

Fig. lift. Frontal view of same specimen. x 1/1. 

Fig. 116. Lateral view of a fragmentary pedicle-valve, showing beak. x 1/1. 

Fig. 12. Martiniopsisl (or ? Spirifer) sp. Ventral view. xl/1. 

Fig. 12ft. Punctate surface of same, much enlarged. 

9. 13-1 3c. Martinia brucei sp. nov. xl/1. 

Fig. 13-/. Surface of same, showing transverse striae, much enlarged. 

Fig. 13e. A larger specimen provisionally referred to Martinia brucei. x 1/1. 

Fig. 14. Martinia sp. a [cf. Spirif era planata Phillips], Ventral view. xl/1. 

Fig. 15. Martinia sp. 6 [aff. rliomboidatis M'Coy]. Ventral view. x 1/1. 

Fig. 15ft. Surface of same, showing longitudinal striae, much enlarged. 

Fig. 16 lihipidomella rnichelini (L'Eveille). Ventral view. xl/1. 

Fig. 17. Schizophoria resupinata (Martin) var. x 1/1. 

Fig. 17«. Surface of same, much enlarged. 

Fig. 18. Sipiamularia sp. ft. Ventral view of a mature specimen. xl/1. 



Trans. Eoy Soc. Edin r - 



Vol. XL VII. 



Plate I— Lee : A Carboniferous Fauna from Nowaja Semlja. 







2a 









26 



2c 






2d 




10 



13 



14 



« 



15a 



10a 



. 




13a 



15 



» * 



20 



21 



106 ^B*^ 



> 



10c 



11 




^^>. 



11a 



136 



13c 







\Sd 



16 




6 



116 




13c 





17a 



196 



17 



£ * 



12 



12a 



19a 



22 



9 



22a 



18 18a 



19 



* <** «* €, 



226 



22c 



22a 1 



<?- v ^•ZA 



A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 185 

Fig. 18a. Squamularia sp. a. Lateral view of same. x 1/1. 

Fig. 18&. ■ Dorsal view of same. x 1/1. 

Fig. 19. Squamularia sp. b. Lateral view of a mature (?) specimen. x 1/1. 

Fig. 19a. Ventral view of same, x 1/1. 

Fig. \9h. Dorsal view of same. x 2. 

Fig. 20. Athyrid. Dorsal view of circular form. x 5/2. 

Fig. 21. Dorsal view of elongate form. x 5/2. 

Fig. 22. Camarospiral sp. Restored outline of largest specimen, slightly enlarged. 

Fig. 22a. Fragmentary pedicle-valve showing the median septum and spondylium. x 3/1 . 

Fig. 22b. Transverse section of a brachial valve, about 2 mm. from umbo. Showing the 

median septum and the absence of a spondylium. x 2/1. 

Fig. 22c. Transverse section of a pedicle valve, about 3 mm. from umbo, x 6/5. 

Fig. 22«\ Transverse section of another pedicle valve, near frontal margin. x 11/7. 



Plate II. 

Fig. 23. Orthotetid, indet. Pedicle-valve. x 1/1. 

Fig. 23a. ■ A brachial valve referred to the same species. x 1/1. 

Fig. 24. Schizophoria sp. Slightly oblique view of a decorticated brachial valve showing the general 
appearance of the muscle-scar. x 1/1. 

Fig. 24a. Isolated pedicle-valve. x 1/1. 

Fig. 24&. ■ Outline of a complete specimen. x 1/1. 

Fig. 25. Chonetes papilionacea (Phillips). Brachial valve. 1/1. 

Fig. 26. Chonetes sp. Coarse-ribbed form. Pedicle-valve, x 1/1. 

Fig. 26a. The same, enlarged. x 2/1. 

Fig. 27. Chonetes sp. Smooth-ribbed form. External view of a pedicle-valve. x 1/1. 

Fig. 27a. Internal view of another pedicle-valve. x 1/1. 

Fig. 28. Chonetes sp. Dorsal view of a specimen, showing the ribbing characteristic of Chonetes cf. 
crassistria [Vaughan]. x 1/1. 

Fig. 28a. The same, enlarged. x 2/1. 

Fig. 29. Productus cf. margaritaceus (Phillips), x 1/1. 

Fig. 30. Productus spinidosus J. Sowerby. Fragmentary pedicle-valve. x 1/1. 

Fig. 31. Productus longispiwis J. Sowerby. Slightly lobate form. x 1/1. 

Fig. 32. var. ? Lateral view of a pedicle-valve belonging to the globose and fine-ribbed 

form, x 1/1. 

Fig. 33. Productus concinnus J. Sowerby. Brachial valve (partly as a mould), x 1/1. 

Fig. 34. Productus (Proboscidella) nysti de Koninck. Ventral view. x 1/1. 

Fig. 34a. Outline of same specimen, showing part of the brachial valve in apposition. 

x 1/1. 

Fig. 35. Productus coatsi sp. nov. The specimen is slightly distorted. x 1/1. 

Fig. 35«. Outline of another specimen. x 1/1. 

Fig. 36. Productus keyserlingianus de Koninck, var. (The ornamentation has been completed from 
another specimen.) x 1/1. 

Fig. 36a. Outline of the same. x 1/1. 

Fig. 37. Productus cherneyensis sp. nov. x 1/1. 

Fig. 37a. Outline of the same. x 1/1. 

Fig. 37 b. A small specimen probably representing the juvenile stage of Productus cherneyensis. 

xl2/5. 

Fig. 38. Productus elegans M'Coy. x 1/1. 

Fig. 38a. A young example. x 1/1. 

Fig. 38b. The same, enlarged. x 2/1. 

Fig. 39. Orthonychia sp. Dorsal view. x 1/1. 



186 A CARBONIFEROUS FAUNA FROM NOWAJA SEMLJA. 

Fig. 'Mia. Orthonychia sp. Lateral view of the same specimen, enlarged. x 2/1. 

Fig. 40. Conocanlium rostratum (Martin). x2/l. 

Fig. 41. Gonocardium ahvforme (J. de C. Sowerby). x 1/1. 

Fig. 4 la. Posterior view of the same. x 1/1. 

Fig. 1-. Murchisonia aff. archiaciana de Koninck. x 7/4. 

Fig. 43 Loxonema sp. b. x 5/2. 

Figs. 44-446. Mourlonia leevis (M'Coy). x 8/5. 

Fig. 45. Palxolima aff. simplex (Phillips). x 2/1. 

Fig. 46. Loxonema Ixvigatum (Eichwald). x 12/7. 

Fig. 47. Turbonitella biserialis (Phillips). Mature example. x 8/5. 

Fig. 47 a. A young example, showing the " semimlcatus" stage. x 5/3. 

Fig. 48. Loxonema sp. a. Outline. x 3/2. 

Fig. 48a. Mature whorl. x 3/1. 

Fig. 49. Loxonemal sp. [Macrocheilinal]. x 6/1. 
Fig. 50. Pygidium of undetermined trilobite. x 8/3. 
Fin. 50a. A segment of the same. x 3/1. 



ns. Roy. Soc. Eclin r - y j XLVII 

Plate II. — Lee: A Carboniferous Fauna from Nowaja Semlja. 




23 




23a 



* 



29 




31 



<• 




39 



37 



► 




24 



246 






30 




32 



34a 




37a 



§ 



39a 



W 



43 



I: 



44 



'P *J 



45 



46 



47 





35 



35a 



# 



376 



24a 



26 




26a 




28 




36 




38 





41 



41a 



40 



^ # 



44a 



446 




47a 48a 



49 



27 





27a 




28a 



36a 




386 



38a 




50 



50a 



( 187 ) 



VIII. — Scottish National Antarctic Expedition : Osteology of Antarctic Seals. By 
Robert B. Thomson, M.B., Ch.B., University of Edinburgh. (With One Plate.) 

(MS. received April 26, 1909. Read July 4, 1909. Issued separately October 12, 1909.) 

Through the courtesy of Dr W. S. Bruce, the leader of the Scottish National 
Antarctic Expedition, and of the late Professor D. J. Cunningham, the skeletons of the 
seals collected in the Antarctic during the voyage of the Scotia have been placed in my 
hands for study, and I wish to express my gratitude to these gentlemen for their great 
kindness in placing all the necessary requirements at my disposal. I also wish to state 
that Dr Bruce has given to the Museum of the University of Edinburgh a choice of 
the various skeletons, which, together with the specimens already obtained, will form 
a comprehensive and almost unique collection of the Antarctic seals. 

The seals obtained during the stay of the Scotia in the Antarctic seas numbered 
forty-four, and embrace all the Phocidse found there, with the exception of the 
elephant seal. Of the forty - four, thirty-four are Weddell seals (Leptonychotes 
Weddelli) ; four are saw-toothed seals (Ogmorrhinus or Lobodon carcinophaga) ; 
three are leopard seals (Stenorliyncus leptonyx) ; one is a sea-lion (Otaria Jubata) ; 
while two are Ross seals (Ommatophoca Rossi). 

A most careful account of date of capture, sex, probable age, full measurements, and 
weight was recorded in a special log-book. In some cases the skeletons were cleaned 
by immersing the carcases beneath the ice, when amphipods and cushion-starfish did 
the work ; in other cases the carcases were conveyed in special tanks and cleaned after 
their arrival home. 

The osteology of the more common species of the Phocidse represented in this collec- 
tion, such as the Weddell seal, saw-tooth seal, sea-lion, and leopard seal, appears to have 
been pretty fully worked up, and especially so in a minute and careful description by 
Principal Sir William Turner ; for the present, therefore, the endeavour of this work is 
to present a detailed description of the Ross seal, as I have not been able to find such 
of the skeleton of this species, with the exception of the skull. In the second place, I 
purpose making a summary of the large collection of other seals with regard to such 
points as dentition and vertebral formula, since the collection is large enough to obtain 
a satisfactory average. 

Ross Seal (Ommatophoca JRossi). 

The first specimen of Ommatophoca was obtained by Sir James Ross in 1840, in 

the pack-ice in the Ross Sea, and is now preserved in the British Museum as the 

type-specimen. I am indebted to the authorities of the British Museum and to 

Dr 0. Thomas for the privilege of examining this specimen for purposes of com- 
TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 8). 28 



188 MR ROBERT B. THOMSON: SCOTTISH NATIONAL ANTARCTIC EXPEDITION: 

parison. Unfortunately, it is not a complete skeleton : both fore-limbs with the 
exception of one scapula, both hind-feet, one dorsal vertebra, and two pieces of sternum 
are a wanting. A description of the skull was made by Dr J. E. Grey in 1844. For 
over fifty years it remained the sole representative of its race in natural history 
collections, as not till 1899 was the Eoss seal again captured by the Belgian Expedi- 
tion (1898-9). Since then numerous specimens have been obtained and the skins 
and crania brought home by the Southern Cross (1898-1900), Discovery (1901-4), 
Morning (1902-4) ; while the Scotia captured two specimens — a male and a female, of 
which the male skeleton is complete with the exception of two segments of the sternum 
and one patella. The female skull has been unfortunately damaged, the occipital 
region being wanting ; but a perfect set of teeth, of so much importance in Ommatophoca, 
is preserved. Further, the whole respiratory apparatus of the male Ross seal has been 
brought home, and is of great interest — as the voice of this animal, which has been 
described by Dr Racovitza, with the peculiar inflation of the pharynx, has attracted 
the notice of, and been recorded by, most Antarctic explorers. 

The method of describing the skeletons is that adopted by Sir William Turner 
in Challenger Reports, and I wish to express my appreciation of its arrangement and 

carefulness. 

Surface Measuremenis {from Log-Book). 

No. 2. No. 43. 

Date of Capture— 6th February 1903. 28th February 1904. 

Sex — Female. Male. 

Age — Adult. Adult. 
Length — 

Nose to tip of tail, 90". 89-5". 
Girth— 

Anterior or neck, 41". 38 - 5". 

Axillary, 52". 55"5". 

Greatest (9 ins. behind axillary), 54". (12 ins. behind axillary), 56*5". 
Tail, 5". 
Fore nippers — 

Outer edge, 16". 17-2". 

Base, 8£". 
Hind nipper — 

Outer edge, 16". 18-4". 

Inner edge, 14£". 18 -7". 

Stretch, 25". 24". 

Base, 107". 11". 

Eye, l£"xl". 1-4" x 8". 

Interorbital space, 6". 6"5". 

Total weight, 450-500 lbs. 400 lbs. 

Remarks. 

Intestines of No. 43 measured from pyloric end of stomach to anus 268". Cuttlefish 
beaks, pieces of cuttlefish, and what appeared to be scales of fish were found in the 
stomach. A great quantity of tape-worms were present in the gut. 



OSTEOLOGY OF ANTARCTIC SEALS. 189 



No. 43 of Collection (Male). 



The skull is characteristically short and broad. The greatest length was from 
premaxilla to the occipital condyles, whilst the greatest breadth was between the two 
zygomatic arches, just at the articulation between its two component parts — the zygo- 
matic process of the temporal and the malar bones. 

A comparison of this skull was made with those of an adult Weddell and sea-leopard 
in regard to length and breadth — an index being framed. Thus, the length-breadth 
index of the Ross skull was 72'4 ; of the Weddell skull, 62*13 ; of the leopard seal, 55-07. 
The nasals measured 79 mm. long, and were completely ankylosed. They articulated 
with the frontal and superior maxillary bones. The part between the two frontals 
amounted to about three-quarters of the entire length of the bones (57 mm.), and was 
triangular in form, while the remaining quarter between the two superior maxillary bones 
was quadrilateral. 

The distance between anterior edge of the outer border of the nasals and the tip of 
premaxilla was 18 mm., the superior maxilla thus forming to this extent part of the 
outer boundaries of the anterior nares. " The length of the section of the boundary 
formed by one of these bones varies in the different specimens from 9 to 1 7 millimetres " 
(Barrett Hamilton, Resultats du Voyage du S.Y. " Belgica," 1897-1899, p. 5). 
The anterior nares sloped downwards and forwards at an angle of 59° with the hard 
palate. In Weddell seal this angle was 48°, in leopard seal 35°. 

The anterior nares were bounded from above downwards by the anterior border of 
the nasal bones, part of nasal borders of superior maxilla, and by the premaxillary bones. 
Looking into the anterior nares, one was struck with two points — the thickness of the 
meso-ethmoid, and the extremely convoluted arrangement of the turbinate bones. The 
anterior edge of the vomer was received between the meso-ethmoid and the premaxillary 
bones. The premaxillary bones supported the two incisor teeth ; their palatal parts were 
triangular in form, and, as before mentioned, their nasal parts did not extend so high as to 
meet the nasal bones. The ante-orbital process of the superior maxilla was well marked, 
and lay in the same transverse plane as the infra-orbital foramen, below which is a 
definite depression from which a groove leads to the orbital floor. The widest part of 
the hard palate was situated well behind the last molar, and was 7 mm. in front of outer 
end of articulation of palatal processes of the superior maxilla and palate bone. It was 
almost flat, showing only a trace of a concavity. The posterior border of the hard 
palate was concave, and possessed a faint post-nasal spine. On the norma verticalis of 
the skull, at the junction of the frontal bones, i.e. at the seat of the anterior fontanelle 
in the young, is situated an opening between the bones. Its margins are irregular and 
spiculated (see fig. 1). As there is no history of an injury, the animal not being 
clubbed, but shot in the thorax, it appears to me to be a persistent anterior fontanelle. 
In the female Ross seal this condition is only faintly represented, but in the plate of 
the Ross seal of the Belgian Expedition a similar well-marked deficiency is to be 



190 MR ROBERT B.THOMSON: SCOTTISH NATIONAL ANTARCTIC EXPEDITION: 

observed. The occipital bone presented occipital crests which were well marked, as also 
were the par-occipital processes. Two well-marked supra-occipital venous foramina 
were to be noticed in the middle line of the occipital squama. The condylar articular 
surfaces were highly convex and approached close to each other on the inferior aspect 
(7 mm.) ; above, they were distant 41 mm. 

The basi-occipital was quadrilateral in form, flat and thick. The tympanic bulla 
was smooth and rounded. The carotid canal was separated from the jugular foramen 
by a thick bar of bone. In the interior of the skull the tentorium cerebelli and falx 
cerebri were partly ossified. 

In the female skull the nasal bones were not ankylosed, and measured 55 mm. long. 
The distance between the anterior edge of the nasals and the premaxilla was 20 mm. on 
right side, 18 mm. on left. 

The Lower Jaw. 

The body was long and somewhat slender, due to small size of teeth, with lower 
border slightly incurved. The ramus formed with the body a moderate angle and 
possessed a well-marked sub-condyloid process. The condyle, which was convex and 
elongated transversely, was separated from the coronoid process by a shallow sigmoid 
notch. The coronoid process was short and pointed. 



Dentition. 

The dentition of the Ross seal is remarkably feeble, and varies more especially with 
regard to the number of post-canines. Dr E. A. Wilson, in vol. ii. of Natural History 
of National Antarctic Expedition, 1907, states : "The food of the Ross seal consists 
mainly of soft-bodied cephalopods, and to this end has developed the incisors and canines 
into needle-pointed recurved hooks of great delicacy and has allowed its post-canines to 
degenerate. The gums presumably can manipulate such food as well as could molar 
teeth, and so we find in some cases the post-canines are small and insignificant, whereas 
in others they are loose and useless, and occasionally absent altogether." He also gives a 
complete table of the dentition of all the Ross crania which have been brought home, 
and which shows the great variations one meets in the dental formula. To this list I 
would add the formulae of the two specimens of the Scotia : — 



No. 2. Female, 


2. 1. 5 
2. 1. 5 ' 


No. 43. Male, 


2. 1. 5 

Oil K 



21 1. bl 



No. 2 skull presents for examination a perfect set of teeth, both upper and lower. 
The upper incisors are more powerful than the lower, but both present, like the canines, 
the well-marked recurved, needle-pointed character. The post-canines are recurved 



OSTEOLOGY OF ANTARCTIC SEALS. 



191 



backwards and three-cusped, the central one being longest. In all but the first the 
post-canines were double-rooted. 

In No. 43 the teeth are not so well developed, and in the upper jaw the left outer 
incisor and canine are broken. The dentition is not quite easily made out. There is 
no fifth post-canine, but small sockets are to be observed ; hence we may presume that 
they were present, having been partly fixed in the natural state by the fleshy gum, a 
condition pointed out by Weber, Bennet, and Wilson {National Antarctic Expedition : 
Natural History, vol. ii. p. 46). 

Again, in the lower jaw no central incisor is present, but the two halves of the jaw 
have been disarticulated. On placing the two halves together, it would seem that no 
room at all is left for central incisors. A comparison of the measurements of the two 
lower and upper jaws between outer side of canines shows : — 

No. 2. Upper diameter, 44 mm. 

Lower ,, 30 ,, 
No. 43. Upper diameter, 48 ,, 

Lower ,, 26 ,, 

In No. 2, which has not been disarticulated, there is a fair amount of fibrous tissue, 
so it is possible to explain the difference in diameters by the absence of this. Hence, 
with the fibrous tissue in place in No. 43, room would be made for central incisors. 

Against this may be stated the fact that the upper central incisors are very feeble, 
and, as the whole of the upper teeth are stronger than the lower, it is possible that the 
lower central incisors may not have developed. 



Measurements of Skulls. 





Eoss Seal, 


Ross Seal, 


Weddell 


Leopard 




No. 43. 


No. 2. 


Seal. 


Seal. 




mm. 


mm. 


mm. 


mm. 


Extreme eondylo-premaxillary length . 


250 




272 


394 


,, inter-zygomatic width .... 


181 




169 


217 


,, width behind external meatus 


167 




177 


198 


Greatest width of palate ..... 


76 


71 


58 


90 


Width between outer side of base of upper canines 


48 


44 


52 


75 


,, ,, ,, lower canines 


26 


30 


30 


50 


Length of palate in line of suture to central incisor 


81 


82 


105 


157 


Height of skull — basion to mid-occipital crest 


83 




84 


110 


Smallest interfrontal diameter .... 


45 


49 


30 


45 


Length of nasals ....... 


79 


55 


61 


103 


Greatest width, anterior nares . . • . 


43 


34 


36 


44 


Length of mandible ...... 


168 


170 


171 


301 


Width between outer end of condyle . 


182 


168 


57 


205 



The vertebral formula was 



L, 



Ud, 



192 MR ROBERT B. THOMSON: SCOTTISH NATIONAL ANTARCTIC EXPEDITION: 

Cervical Vertebral. 

The bodies of the posterior six cervical vertebrae were keeled on their ventral aspect, 
the keel being more pronounced in the anterior than in the posterior series. On con- 
trasting the bodies of the Ross seal with those of the Weddell and leopard seal, a 
great difference was noticed with regard to their length and breadth. If the 4th 
cervical be taken as a typical vertebra, we find the length index of the body to be : — 

In the Ross seal — Length 33 48 x 100 



Breadth 48 33 



= Index 145 "4. 



In leopard seal — Length 51 47xl00 = t , 99-16 
Breadth 47 51 



In Weddell seal— Length 38 
Breadth 38 



Index 100. 



If these indices be compared with the length-breadth index of the skull, one is 
justified in saying that the short and broad character of the Ross skull can also be 
applied to the bodies of the vertebrae. The neural ring was triangular. A depth- 
breadth index shows much the same difference as the bodies in the three species 
contrasted. In the Ross seal, the index was 35 1 ; in Weddell seal, 54*5 ; and in the 
leopard seal, 50. 

The transverse processes in all except the 7th were perforated at their roots by a 
vertebraterial foramen, and possessed at their extremities well-marked tubercles in all 
except the 2nd and 7 th. The spinous processes exhibited a marked difference in the 
various vertebrae. Thus the axis possessed a massive spine with four well-marked 
tubercles — two anteriorly and two posteriorly. The 7th had a well-marked spine, not 
bifid ; the 3rd and 4th had only a trace of a spine, while the 1st possessed a well-marked 
tubercle representing a spine. The 5th and 6th presented fairly prominent spines, bifid 
— a character not shown by either the Weddell or leopard seals. The type-specimen 
in the British Museum also exhibits this character. The anterior articular processes 
looked upwards and inwards, the posterior downwards and slightly forwards. 

The atlas has powerful lateral masses supporting anteriorly deeply concave articular 
surfaces for occipital condyles, while the posterior articular facets for the axis were flat. 
The transverse processes were plate-like, and projected outwards and downwards. The 
lamina was perforated on each side by a foramen for the vertebral artery, which foramen 
was situated internal to the upper ends of the articular facets for the occipital condyles. 

The axis did not present so well-marked an odontoid process as either the Weddell 
or leopard seal. On the inferior aspect of the odontoid process was a triangular facet 
for articulation with the atlas, whilst superiorly it had a small oval facet for articulation 
with the transverse ligament. The transverse processes were feeble and pointed, whilst 
their superior roots were mere spicules of bone. The ventral aspect of the body was 
markedly keeled. 



OSTEOLOGY OF ANTARCTIC SEALS. 193 



Dorsal Vertebras. 



The bodies of the anterior four were keeled, the keel becoming less pronounced as 
we proceeded from 1 to 4. At the posterior end of the dorsal region, 14 and 15 pre- 
sented a slight keel, more accentuated in 15 than in 14. The side of the 1st body had 
one and a half facets ; 2 to 10 inclusive (two half-facets, 11 to 14 inclusive) a whole facet 
for the corresponding ribs. The 1 5th dorsal vertebra showed a peculiar arrangement ; thus, 
on the right side of its body a well-marked facet was present for articulation with the 15th 
rib of that side, while on the left a well-marked plate-like costal process, representing 
the last rib, projected from the junction of the body and pedicle outwards, forwards, and 
slightly downwards. It measures 42 mm. in length and 25 mm. in breadth. (See 

fig- 5.) 

The spines of the dorsal vertebras were relatively feeble compared with those of the 
Weddell and leopard seals. They were low, and projected upwards and backwards as 
far as No. 11 ; while 12, 13, 14, 15 projected upwards. 

The transverse processes of 1 to 10 were prominent and presented rounded facets for 
articulation with the tubercles of the ribs. In 11, 12, and 13 the transverse processes 
were feeble but possessed facets. In 14 and right side of 15 they were rudimentary 
and non-articular, while the character of the transverse process of left side of 15 has 
already been noticed. From 9 to 15 the mammillary processes were well defined, while 
accessory processes were faintly marked in 9 to 14. 

Lumbar Vertebras and Sacrum. 

Each body of the lumbar vertebrae was elongated antero-posteriorly and was keeled 
on its ventral aspect. The spines were broad and more pronounced than in either the 
Weddell or leopard seals. The mammillary processes were rounded and directed 
forwards and outwards. The transverse processes were thin and flat, and curved out- 
wards, forwards, and downwards. The sacrum was composed of three segments fused 
together. The ventral surface in its anterior part was broad and flat, with a faint keel, 
while its posterior part, corresponding to the 2nd and 3rd segments, was protuberant 
in the centre, concave laterally, where it presented the openings of the two anterior 
sacral foramina. The posterior surface showed mesially three spinous processes, while 
the lamina? of the three segments were only fused laterally. The fused articular 
processes were prominent at the junction of S x and S 2 , less so at S 2 and S 3 . The 
direction of the 1st posterior sacral foramen was horizontally outwards, in the 2nd and 
3rd upwards and only slightly outwards. Laterally the sacrum presented a definite 
U-shaped articular surface for ilium, while the interval between the two limbs of the 
U and the area posterior to it were hollow for ligamentous attachment. The articular 
surface for the ilium was confined entirely to the first sacral segment. 



194 MR ROBERT B. THOMSON: SCOTTISH NATIONAL ANTARCTIC EXPEDITION: 

Sacral Index. 
Ross seal ..... 108 
Weddell seal . . . . .102-9 

Leopard seal ..... 114'8 

The caudal vertebrae were ten in number. The posterior six were represented by 
elongated bodies alone — the last being cartilaginous. The 1st possessed, in addition to 
the body, a definite neural arch terminating above in a rudimentary spinous process. 
The laminae in the 2nd did not unite superiorly, so that the neural groove was not con- 
verted into an arch. The third possessed on each side of its body rudimentary laminae. 
The bodies of the first four caudal vertebrae were keeled on their ventral aspect. The 
sternum was incomplete, two segments being awanting. Their articulations could not 
be made out definitely, so I refrain from describing them. 

Ribs. 

The ribs were fifteen in number on the right side and fourteen on the left, this anomaly 
being explained in connection with the thoracic vertebrae. The 1st, 12th, 13th, 14th, 
and 15th ribs possessed a single facet each on their heads for articulation with the 
corresponding vertebrae. The 2nd to 11th inclusive possessed two facets separated by 
a distinct ridge. The 1st to 12th ribs inclusive possessed well-defined necks, while the 
1st to 13th showed an oval concavo-convex facet for articulation with the corresponding 
transverse processes. In the anterior series of ribs was a slight depression, in some a 
roughness, on each side of the articular surface of the head, for attachment of ligamentum 
conjugale costarum described by Sir William Turner {Challenger Reports, vol. xxvi., 
Report on Seals, p. 14). The shafts of the ribs were slightly curved, the most pro- 
nounced curvatures being observed in the 2nd, 3rd, and 4th, while the greatest in 
length were the 8th and 9th. The last rib on the right side possessed a head with an 
articular facet, no neck, and a feeble shaft. Its length was 127 mm. 

The Fore- Limb. 

The scapula presented the well-marked sickle shape so characteristic in the Phocidae. 
The anterior and superior or vertebral borders were convex and thin, the inferior or axil- 
lary was concave and rounded. On the axillary border at junction of lower and middle 
thirds in the Weddell and leopard seals was a prominent ridge for muscular attachment. 
No such ridge was present in the Ross scapulae. The dorsum scapulae was divided into 
a pre-spinous and post-spinous area by a feeble spine. This spine was relatively longer 
than in the Weddell and leopard seals, but not so well developed. The pre-spinous area 
was concavo-convex, the post-spinous was flat. There was no acromion process, but in 
the Weddell and leopard seals this was present although feeble. The coracoid process 
was rudimentary and projected forwards, measuring from the bottom of the shallow 






OSTEOLOGY OF ANTARCTIC SEALS. 195 

notch at its base 18 mm. The ventral surface showed a slight subscapular angle cor- 
responding in position to the attachment of the spinous process on the dorsal surface. 
Two faint muscular ridges were present, directed towards the glenoid fossa. The 
glenoid fossa was pear-shaped, deeply concave, with a prominent margin. 

Index. 

Ross seal . . . . .85 

Weddell seal ..... 82-14 

In the leopard seal the epiphysis of the posterior angle was wanting, hence the 
index could not be ascertained. 

The humerus measured 115 mm. long. The head was convex, and its articular 
surface was continued forwards on to the base of the lesser tuberosity, which was 
equally as prominent as the greater one. The shaft was slightly twisted, and pre- 
sented a strongly marked deltoid ridge. The bicipital groove was in the form of a 
shallow groove. The lower end presented a well-marked trochlea for the ulna, a slight 
coronoid fossa, but no radial or olecranon fossae. The capitellum was circular and 
convex. The internal condyle was more pronounced that the external, the external 
supra-con dyloid ridge more so than the internal, which did not possess a supra - 
condyloid foramen or even a process. 

Bones of Forearm. 

In neither were the distal epiphyses fused with the shaft. The ulna measured 
152 mm. and was expanded above from before backwards so as to form a long but 
narrow olecranon process. A coronoid process was not present. The articular surface 
for humerus was concavo-convex, while the facet for radius above was flat and con- 
tinuous with that for the humerus. The shaft was flat, with a rounded anterior border 
and a sharp posterior border. The lower end was conical and articulated with radius 
slightly, and with cuneiform and pisiform. The radius was 135 mm. long. The head 
possessed a deep cup-shaped cavity for articulation with the capitellum of humerus, 
while the margin of the head was non-articular except on its inner side, where there 
was a well-marked facet for articulation with the ulna. On the neck a faint tubercle 
was noticed, while the shaft was paddle -shaped, with a definite roughness on its outer 
aspect for pronator radii teres. The lower end was grooved on its posterior aspect, and 
articulated with the ulna and scapholunar. 

Humero-Radial Index. 

Length of radius x 100 
Length of humerus 

Ross Seal. Sea Leopard. Weddell Seal. 

117-4 107-1 98-1 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART I. (NO. 8). 29 



19G MR ROBERT B. THOMSON: SCOTTISH NATIONAL ANTARCTIC EXPEDITION: 

Carpal Bones. 

These were seven in number, the scaphoid and semilunar being fused together to 
form a single bone. The carpal bones were devoid of ridges or processes. The articu- 
lations were difficult to make out, but I have referred these as follows : — 

The scapholunar articulated with radius, trapezium, trapezoid, os magnum, and 
unciform. 

The cuneiform articulated with the ulna, pisiform, and unciform. 

The pisiform was a mere nodule, and articulated with the cuneiform and ulna. 

The trapezium articulated with scapholunar, trapezoid, and 1st and 2nd metacarpals. 

The trapezoid articulated with scapholunar, trapezium, os magnum, and 2nd meta- 
carpals. 

The os magnum was small, and articulated with the 2nd, 3rd, and 4th metacarpals, 
and with the trapezoid, scapholunar, and unciform. 

The unciform articulated with the scapholunar, cuneiform, and os magnum, and with 
the 4 th and 5 th metacarpals. 

Metacarpal Bones. 

The 1st metacarpal was the longest, and possessed both a proximal and a distal 
epiphysis not united with the shaft. The 2nd was longer than the 3rd, 4th, or 5th, 
which were of about equal length, and the epiphysis for the head of each was separate 
from the shaft. 

Tlie Hind-Limb. 

The innominate bone, measuring in length 215 mm., consisted of three parts — ilium, 
ischium, and pubis. The ilium was short and stunted. The ischium and pubis were 
long, somewhat narrow, and enclosed between them a very large obturator foramen — 
larger indeed than the ilium itself. 

The acetabulum was circular in form, entirely articular except an elongated, hollow 
area at the bottom with a definite margin. The brim of the acetabulum was deficient 
in its lower aspect, thus forming a cotyloid notch. The ilium measured from centre 
of acetabulum to the highest part of the iliac crest 74 mm. The length of its crest was 
102 mm. Its dorsal and ventral aspects were flattened and did not possess any definite 
muscular ridges. At the posterior part of the ventral surface was a U -shaped articular 
surface for that part of sacrum corresponding to its first segment, while there was a 
roughened area for the attachment of the posterior sacro-iliac ligament. 

The os pubis projected backwards and slightly downwards from the acetabulum, 
and its junction with the ilium was marked by a prominent pectineal eminence. From 
this tubercle to the symphysis, which was small, the pubis measured 130 mm. 

The ischium passed backwards and slightly upwards from the acetabulum, and 
possessed a pointed spine. 

The femur measured 114 mm. long. The head was entirely articular except in its 
postero-superior quadrant, where there was a non-articular area continuous with the 



OSTEOLOGY OF ANTARCTIC SEALS. 197 

non-articular neck for attachment of the ligamentum teres. Dr Keith, in Human 
Embryology and Morphology, p. 385, states: "The ligamentum teres is isolated 
from the capsule by the development of the head of the femur, which expands as a wing 
on each side of the ligamentum teres, and by fusion of the wings isolates it from the 
capsule." In the Ross seal femur the wings have not yet met ; hence, according to 
Keith's view, the ligamentum teres would still be in connection with the capsule. (See 
fig. 6.) 

The trochanter major was a broad, rough prominence projecting from the outer side of 
the junction of the neck of the bone with the shaft. There was no trace of a trochanter 
minor, as is to be found in Otaria Jubata ; neither was there a third trochanter. 

The shaft, flat and expanded, possessed an anterior and a posterior surface, both 
being very slightly rounded and devoid of muscular ridges. The outer border was 
short and concave, the inner convex and very rough for muscular attachment. 

At the posterior end of the bone, the trochlear surface for the patella was flattened 
and ascended as high on its outer as on the inner side. It was quite distinct from the 
condylar surface, but closely approximated to the external one, being separated by a 
very narrow groove. 

The internal condylar surface was flat and circular, the outer one oval and slightly 
convex. 

The intercondyloid fossa was shallow. 

The patella was 34 mm. long and 24*5 mm. broad. It was oval in form, and its 
articular surface was slightly concave, and did not exhibit any subdivision into special 
articular areas. Its anterior end was much thicker than the posterior. 

The tibia and fibula were fused together at their upper ends. The tibia measured 
247 '5 mm. Its upper end was elongated transversely and possessed two articular areas 
for the condyles of the femur, separated by a rough intermediate area. The shaft was 
slightly curved, fiat and smooth on its ventral aspect, deeply concave on its dorsal 
aspect for origin of the tibialis posticus muscle. The nutrient foramen was a large one. 
The inner border was broad and rough, the outer one thin and rounded. The lower end 
was not united with the shaft, and from it projected a slight malleolus, which was non- 
articular. The ventral and dorsal aspects of the lower end were deeply grooved, while 
the articular area for astragalus was reniform and concave. 

The fibula was 233 mm. in length. The upper end was fused with the tibia, whilst 
the lower end was grooved on the outer side by the peroneal tendons, and had on its 
inner aspect an oval facet for outer surface of astragalus. The shaft was straight, 
and its surface broad for muscular origins. The lower epiphysis was not united to 

the shaft. 

Tibia- Femoral Index. 





Length of tibia x 100 






Length of femur 




oss Seal. 


Weddell Seal. 


Leopard Seal. 


217-1 


203-9 


2135 



198 MR ROBERT B. THOMSON: SCOTTISH NATIONAL ANTARCTIC EXPEDITION: 

Tarsal Bones. 

The astragalus was the largest of the tarsal bones. It consisted of a body and two 
large processes. The posterior process projected backwards beyond the hinder end of 
os calcis, and presented a well-marked groove on its under aspect. Its anterior process 
or head passed as far forward as did the os calcis, and articulated with the scaphoid, 
cuboid, and fore-part of os calcis. The superior surface of the body presented a quadri- 
lateral convex surface for articulation with the lower end of tibia, continuous externally 
with a triangular facet for fibula. Its internal surface did not articulate with the 
internal malleolus. The inferior surface articulated with os calcis and presented a deep 
groove for an interosseous ligament. 

The os calcis possessed an attenuated posterior or peroneal process, and a feeble 
internal or sustentacular one. It articulated with the astragalus and fibula above, with 
the cuboid in front, and was grooved on its outer aspect by the peroneal tendons. 

The cuboid presented on its inferior aspect a well-marked plantar tubercle and a 
deep peroneal groove. It articulated with the os calcis, astragalus, scaphoid, external 
cuneiform, and 4th and 5th metatarsal bones. 

The scaphoid articulated with the astragalus, cuboid, and the three cuneiforms. Its 
posterior surface was not entirely concave, being convex in its outer part. 

The internal cuneiform articulated with the scaphoid, middle cuneiform, and 1st and 
2nd metatarsals. 

The middle cuneiform was the smallest of the three, and was completely shut out of 
the plantar aspect of the foot. It articulated with the cuneiform on either side of it, 
with scaphoid behind and 2nd metatarsal in front. 

The external cuneiform articulated with the scaphoid, cuboid, 2nd and 3rd 
metatarsals. 

Metatarsal Bones. 

The 1st was the longest, the 5th next in order, then 4th, 2nd, and 3rd, which was 
the shortest. 

None of the epiphyses were fused with the shafts. The base of the 2nd showed the 
characteristic hook-shaped base. 

The articulations of their bases were as follows : — 

1st with internal cuneiform and 2nd metatarsal, 
2nd with the three cuneiforms 1st and 3rd metatarsal, 
3rd with external cuneiform 2nd and 4th ,, 
4 th with cuboid 3rd and 5th ,, 

5th with cuboid 4th ,, 

while its base projected outwards as an elongated process. 

The terminal phalanges of both fore and hind limbs were retained in connection with 
the skins. 



OSTEOLOGY OF ANTARCTIC SEALS. 199 

Summary. 

The skulls correspond in their distinguishing characters from other Antarctic seals 
to those tabulated by Sir William Turner in the Challenger Reports (Voyage of 
" Challenger" Eeports on Seals, p. 66). 

With regard to the rest of the skeleton, one specimen alone does not provide 
altogether sufficient evidence ; still, I note the following points which may be of interest 
on further skeletons being brought home and examined : — 

1. Much greater length-breadth index of vertebrae. 

2. Bifid character of 5th and 6th cervical spines, also present in type-specimen in 

British Museum. 

3. Sacral index. 

4. Scapular index. 

5. Humero-radial index. 

6. Non-articular area for ligamentum teres on head of femur continuous with non- 

articular neck. 

7. Tibio-femoral index. 

Number of Vertebra in Phocid^e. 

In the Resultats du Voyage du S.Y. " Belgica" : Zoologie, p. 7, Mr Barrett 
Hamilton states that in most seals the numbers of the thoracic and lumbar vertebras 
appear to be usually 15 and 3 (5?), more rarely 14 and 6. In the Antarctic seals 
without exception I have found the vertebral formula in the adult to be C 7 D ]5 L 5 
S 3 Cd 10 or u . 

With a view to obtaining as large a number of specimens as possible, in order to 
form an average, I have made a careful examination of the seal skeletons in the 
University Museum here, the Museum of Science and Art of Edinburgh, British 
Museum of London, and give a table of the vertebral formula in 47 specimens : — 

Weddell seal : 27 with formula C 7 D 15 L 5 S 3 Cd 10 or n . 
Crab-eating seal : 2 with formula C 7 D 15 L 5 S 3 Cd ]0 . 
Sea-leopard : 2 with formula C 7 D 15 L 5 S 3 Cd 10 . 
1 ,, ,, C 7 D 14 L 6 S 3 Cd n . 
Ross' seal : 1 with formula C 7 D 15 L 5 S 3 Cd 10 . 
Sea-lion : 2 with formula C 7 D 15 L 5 S 3 Cd 10 . 
Fur seal : 1 at C 7 D 15 L 5 S 3 Cd n . 
Crested seal : 1 at C 7 D 15 L 5 S 3 Cd 10 . 
Common seal : 2 at C 7 D 15 L 5 S 3 Cd n . 
Elephant seal : 1 at C 7 D 15 L 5 S 3 Cd 10 . 
New Zealand fur seal : 1 at C 7 D 15 L 5 S 3 Cd 10 . 
West Indian seal : 1 at C 7 D 15 L 5 S 3 Cd 12 . 



200 MR ROBERT B. THOMSON : SCOTTISH NATIONAL ANTARCTIC EXPEDITION : 

Grey seal : 1 at C 7 D 15 L 5 S 3 Cd 12 . 

Bearded seal : 1 at C 7 D 15 L 5 S 3 Cd n . 

Northern fur seal : 1 at C 7 D 15 L 5 S 3 Cd 10 . 

Altogether 46 specimens with formula of C 7 D 15 L 5 S 3 Cd 10 to 12 . 

One specimen with formula of C 7 D u L 6 S 3 Cd n . 

The number of caudal vertebrae seems to vary from 10 to 12, while in some 
cases one could not be positive, as frequently the terminal vertebrae are preserved 
in the skins. 

SUPRA-CONDYLOID FORAMEN. 

With a view to the formation of a percentage of such foramina found in the humeri 
of the Phocidse, I have examined altogether 100 humeri, made up as follows : — 

60 Weddell seals. 

4 crab-eating seals. 

6 leopard seals. 

4 sea-lions. 

2 Ross seals. 

4 elephant seals. 

4 South American fur seals. 

2 New Zealand fur seals. 

2 West Indian seals. 

2 common seals. 

2 North American fur seals. 

Total, 92 

In these 92 no foramen was present, but in 4 Weddell humeri there were small 
supra-condyloid processes. 

In the following 8 humeri, supra-condyloid foramina were present : — 

2 grey seals. 

2 bearded seals. 

2 crested seals. 

2 common seals. 

Total, 8 
thus giving a percentage of 8. 

Summary of Dentition. 

As the dentition of the Phocidse is of some importance, especially with regard to 
variation, I have formed a table of the dentition of the seals under my observation. I 



OSTEOLOGY OF ANTARCTIC SEALS. 



201 



have also examined the crania of the seals in the University Museum of Science and 
Arts, Edinburgh, and include such in the following list : — 

2. 1. 5 



Weddell seal : 2 1 specimens with formula . 
1 specimen ,, ,, 



2. 1. 5 

2. 1. 6 
2. 1. 5 



The fourth post-canine in the upper jaw was obviously split into two, thus account- 
ing for the variation. 



Sea-leopard : 5 specimens with dental formula 
Crab-eating seal : 7 specimens with formula 
Sea-lion (Otaria Jubata) : 2 specimens with formula 
1 specimen „ „ 



2. 1. 5 

2. 1. 5 

2. 1. 5 
27T~5 

3. 1. 5 
3. 1. 5 

3. 1. 6 

3. 1. 5 



LITERATURE. 



The Zoology of the Voyage of the " Erebus " and " Terror" pp. 7-8, plates vii. and viii. 

J. E. Gray, Ommatophoca Rossi, 1844. 

Challenger Reports, vol. xxvi., Sir Wm. Turner, 1888. 

Resultats du Voyage du S.Y. " Belgica," G. E. H. Barrett Hamilton, 1901. 

National Antarctic Expedition : Natural History, vol. ii., E. A. Wilson. 



DESCRIPTION OF PLATE. 
Fig. 1. — Ross skull, No. 43. Superior aspect. 

Fig. 2. — Ross skull, No. 43. Lateral aspect. 

Fig. 3. — Ross skull, No. 43. Anterior aspect. 

Fig. 4. — Ross skull, No. 43. Inferior aspect. 



Fig. 5. — Ross seal, No. 43. 15th dorsal vertebra. 
Showing costal process on right side, and 
facet for rib on left side. 



Fig. 6. — Eoss seal, No. 43. Femur showing 
non-articular area on head. 




18 FEB. 1910 



Trans. Roy. Soc. Edin. Vol. XLVII. 

Robert B. Thomson — " Scottish National Antarctic Expedition : Osteology of Antarctic Seals." 



-ir#*r»*- 





FlG. 1. 



Fig. 2. 




mm 




Fig. 3. 



Fig. 4. 





Fig. 5. 



Fig. 6. 



The Transactions of the Royal Society of Edinburgh will in future be Sold at the 

following reduced Prices : — 





Price 


Price 






Price 


Price 






Price 


Price 


Vol. 


to the 


to 




Vol. 




to til 


e 


to 




Vol. 




to the 


to 




Public. 


Fellows. 




Public. 


Fellows. 




J 


Public. 


Fellows. 




£ s. 


d. 


£ s. 


d. 




£ 


s. 


d. 


£ s. 


d. 




t 


s. 


d. 


£ s. d. 


VI 


11 


6 


9 


6 


XXXIII Pt. 1 


1 


1 





16 





XIII 


2 


2 





1 11 


VII. 


18 





15 





„ Pt. 2 


2 


2 





1 11 





XLIII. 


2 


2 





1 11 


VIII. 


17 





14 





„ Pt. 3 





12 





9 


6 


XIIV. 


Not yet p 


ublished. 


IX. 


1 





17 





XXXIV. 


2 


2 





1 11 





XIV. Pt. 1 


1 


9 





1 2 


X. 


19 





16 





XXXV. *Pt. 1 


2 


2 





1 11 





„ Pt- 2 


1 


7 





1 


XI. 


14 


6 


12 





Pt- 2 


1 


11 





1 3 


6 


„ Pt. 3 


I 


13 


9 


1 5 3 


XII 


14 


6 


12 





Pt. 3 


2 


2 





1 11 





,, Pt. 4 





4 


6 


3 6 


XIII. 


18 





15 





Pt. 4 


1 


1 





16 





XIVI Pt. 1 


1 


1 


10 


16 6 


XIV. 


1 5 





1 1 





XXXVI. Pt. 1 


1 


1 





16 





„ Pt. 2 


1 


5 


8 


19 4 


XV 


1 11 





1 6 





Pt. 2 


1 


16 


6 


1 7 


6 


„ Pt. 3 


1 


7 


3 


1 11 


XX. Pt. 1 


18 





14 





Pt. 3 


1 








16 





XIVI I. Pt. 1 





19 


9 


15 


XXII. Pt. 2 


10 





7 


6 


XXXVII Pt. 1 


1 


14 


6 


1 5 


6 












„ Pt. 3 


1 5 





1 1 





„ Pt. 2 


1 


1 





16 















XXVII Pt. 1 


16 





12 





Pt. 3 





16 





12 















„ Pt. 2 


6 





4 


6 


Pt. 4 





7 


6 


5 


8 












„ Pt. 4 


1 





16 





XXXVIII Pt. 1 


2 








1 10 















XXVIII. Pt. 1 


1 5 





1 1 





Pt. 2 


1 


5 





19 















„ Pt. 2 


1 5 





1 1 





Pt. 3 


1 


10 





1 3 















„ Pt. 3 


18 





13 


6 


Pt. 4 





7 


6 


5 


8 












XXIX. Pt. 1 


1 12 





1 6 





XXXIX. Pt. 1 


1 


10 





1 3 















„ Pt. 2 


16 





12 





Pt. 2 





19 





14 


6 












XXX. Pt. 1 


1 12 





1 6 





Pt. 3 


2 


3 





1 11 















„ Pt. 2 


16 





12 





Pt. 4 





9 





7 















„ Pt. 3 


5 





4 





XL. Pt. 1 


1 


5 





19 















„ Pt. 4 


7 


6 


5 


8 


„ Pt.2 


1 


12 


6 


1 5 


6 












XXXI. 


4 4 





3 3 





„ Pt. 3 


1 


6 





19 


6 












XXXII Pt. 1 


1 





16 





„ Pt.4 


1 








16 















Pt. 2 


18 





13 


6 


XLI Pt. 1 


1 


1 





15 


9 












„ Pt. 3 


2 10 





1 17 


6 


„ Pt.2 


1 


9 


6 


1 2 















„ Pt.4 


5 





4 





„ Pt. 3 


2 


5 





1 13 


6 













* Vol. XXXV., and those which follow, may be had in Numbers, each Number containing a complete Paper. 

January 1909. — Volumes or parts of volumes not mentioned in the above list are not for the present on sale 
to the public. Fellows or others who may specially desire to obtain them must apply direct to the Society. As 
the Society reprints from time to time parts of its publications which have become scarce, the absolute correctness 
of this list cannot be guaranteed beyond this date. 



The Papers published in this Part of the Transactions may be had separately, on application to the Publisher*, 



at the following prices , 



No. I., . 
No. II., . 
No. Ill, 
No. IV., 



Price to Public. 


Price to Fellows. 




Price to Public. 


Price to Fellows 


Is. 3d. 


Is. 


No. V., 


4s. 


3s. 


4s. 6d. 


3s. 6d. 


No. VI, 


6d. 


4*4 


Is. 


9d. 


No. VII, 


4s. 


3s. 


3s. 


2s. 3d. 


No. VIII., . 


Is. 6d. 


Is. l|d. 



2 4 MAY 1910 

TRANSACTIONS 



OF THE 



ROYAL SOCIETY OF EDINBURGH. 

VOLUME XLVIL PART II.— SESSION 1909-10. 



CONTENTS. 

No. 





Page 



IX. Lower Paleozoic Hyolithidx from Girvan. By F. R. Cowper Reed, M.A., F.G.S. (With 

Three Plates), .......... 203 

(Issued October 13, 1909.) 

^ X. Report upon the Anatomy and Embryology of the Penguins collected by the Scottish National 
Antarctic Expedition, comprising: (1) Some Features in the Anatomy of the Penguin; 
(2) Tlie Embryology of the Penguin : a Study in Embryonic Regression and Progression. 
By D. Waterston, M.A., M.D., F.R.C.S.E., and A. Campbell Geddes, M.D. (With 
Three Plates), . . ....... 223 

(Issued October 21, 1909.) 

XL The Pharmacological Action of Harmaline. By James A. Gunn, M.A., M.D., D.Sc, . 245 

(Issued December 31, 1909.) 

XII. The Pharmacological Action of Protocatechyl-Tropeine. By Professor C. R. Marshall, . 273 

(Issued January 7, 1910.) 

XIII. The Pharmacological Action of Tutu, the Toot Plant of New Zealand. By Professor 

C. R. Marshall, ......... 287 

(Issued January 29, 1910.) 

XIV. On the Life-history of Hydrobius fuscipes, L. By Frank Balfour-Browne, M.A. (Oxon.), 

F.Z.S. (With Three Plates), . . . . . .317 

(Issued February 1, 1910.) 

XV. Strophanthus sarmentosus : its Pharmacological Action and its Use as an Arrow-Poison. By 

Sir Thomas R. Fraser, M.D., F.R.SS. L. & E., Professor of Materia Medica in the 
University of Edinburgh, and Alister T. Mackenzie, M.A., M.B., Ch.B., Carnegie 
Research Scholar. (With Eleven Plates), . . . . .341 

(Issued February 8, 1910.) 



EDINBURGH: 

PUBLISHED BY ROBERT GRANT & SON, 107 PRINCES STREET, 

AND WILLIAMS & NORGATE, 14 HENRIETTA STREET, CO VENT GARDEN, LONDON. 

MDCCCCX. 
Price Twenty-three Shillings. 



( 203 ) 



IX. — Lower Palaeozoic Hyolithidae from Girvan. By F. R. Oowper Reed, M.A., 
F.G.S. Communicated by Dr Horne, F.R.S. (With Three Plates.) 

(MS. received March 5, 1909. Read June 7, 1909. Issued separately October 13, 1909.) 

Introduction. 

The extensive series of specimens of members of the Hyoliihidce which Mrs 
Elizabeth Gray has collected from the Ordovician and Silurian beds of the Girvan 
district, and submitted to me for investigation, forms the basis of the following memoir. 

Through the kindness of Dr John Horne and Professor J. W. Gregory, special 
facilities have been afforded me of examining the other Scotch examples in the 
Geological Survey Collection, Edinburgh, and the Museum of Glasgow University. 
My thanks are also due to Dr F. L. Kitchin for the opportunity of studying the 
specimens in the Jermyn Street Museum. 

The collections in the Sedgwick Museum, Cambridge, from various localities and 
many Lower Pala30zoic horizons in Great Britain and Ireland, have enabled me to 
compare the Girvan species with previously described and typical forms. The 
specimens in the same museum from Bohemia, named by Barrande himself, have 
proved of much value. Examples in the British Museum have also been consulted. 

The classification of the Hyolithidce drawn up by Holm is here followed, the 
genus Hyolithes being divided into two principal subgenera, Hyolithes sens. str. 
and Orthotheca, but Novak's Ceratotheca is retained as a third independent group of 
generic rank for two doubtful species. The genus Pterotheca is also included in 
the same family, but the reasons for doubting the accuracy of this customary 
position are mentioned and briefly discussed. 

Genus Hyolithes, Eichwald, 1840. 

Subgenus Hyolithes, Eichwald -Holm, 1893. 

Hyolithes ardmillanensis, sp. nov. (PI. I. figs. 1-6.) 

Shell straight, short, broad, rapidly tapering at about 1 in 2|-, dorso-ventrally 
compressed. Ventral face moderately convex, rounded or slightly subangular Dorsal 
face flat. Mouth oblique, with strongly arched dorsal lip. Lateral angles sharp, 
acute. Ventral face ornamented with 18-20 strong, rounded, straight longitudinal 
raised lines, of equal or subequal size, and separated by wider interspaces. Dorsal 
face, with fine concentric arched striae and traces of longitudinal folds, and occasionally 
with shallow submarginal grooves visible in the cast. 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART II. (NO. 9). 30 



I. 


II. 


26-0 


c. 23 - mm, 


10-5 


c. 9-0 „ 



204 MR F. R. COWPER REED ON 

Dimensions :— 

Length ...... 

Width at mouth ..... 

Horizon and Locality. — Balclatchie Group, Ardmillan. 

Remarks. — This is a well-characterised form. Generally only the dorsal face is 
preserved or the internal cast of the shell. The strong, regular longitudinal lineation 
of the ventral face and the short, broad shape of the shell are distinguishing features. 

Holm's H. cymbium* from the Lituites Limestone, appears to be closely allied in 
both these features, but has finer longitudinal lines and a slightly curved shell. 

The species described by REMELE + as H. incequistriatus, from the boulders of 
Orthoceras Limestone in the North German Drift, closely agrees with our species in 
shape and rate of tapering, and the ornamentation is of the same type, though the 
longitudinal lines vary in strength. 

Some unnamed specimens from the Bala beds of Horderley, in the Jermyn Street 
Museum, bear considerable resemblance to H. ardmillanensis in cross-section, rate of 
tapering, and ornamentation, but are too poorly preserved to warrant identification 
with it. 

An operculum (PI. I. fig. 6) occurs associated with the above shells at Ardmillan 
which is of the requisite proportions and may probably be referred to this species. 
It is subelliptical in shape, folded sharply along a line near the ventral margin, 
parallel to the longer diameter, and is conical and moderately elevated, with the apex 
situated on the fold close to the ventral margin. The ventral side is short, narrow, 
very steeply inclined, and almost at right angles to the rest of the surface. The 
general surface of the cone, though rising to the apex, is somewhat concave, and is 
marked with three strong radiating grooves on each side, diverging from the apex, and 
not reaching the margin. As we are dealing with an internal cast, these grooves 
really represent internal ridges in the operculum, as Barrande pointed out. 

Fine radiating lines occur on each side of the sharp, angular junction of the 
ventral side with the general surface, and are traceable (though less distinctly) near the 
margin on other parts of the surface, but over the greater part only very delicate 
concentric striae can be detected. 

The transverse diameter of one such operculum is 11*5 mm. 

Barrande | has figured from Stage Ddl a somewhat similar isolated operculum 
with the internal cast showing three radiating grooves on each side ; these he interprets 
as muscular impressions. 

It should be noted that none of the Girvan opercula have been found in position, 
but their relations to the shell are inferred from the Bohemian species described by 

* Holm, "Sver. Kamb.-silur. Hyol. o. Conul." (Sver. Geol. UndersoJcn., Ser. C, No. 112, 1893), p. 82, t. iii. 
figs. Oil. 

+ Kkmixi', Z,;t,,hr. dent. geol. Gesell, Bd. xl., 1888, p. 670, t. xxviii. figs. 4-6. 
J BaBEAOTJE, Sijst. SiJm: Boheme, vol. iii., 1867, p. 98, pi. ix. figs. 16, 17. 



LOWER PALAEOZOIC HYOLITHID^E FROM GIRVAN. 205 

Barrande with the operculum in place. In these the apex of the cone is always 
nearer the ventral side of the shell ; but Salter regarded it as nearer the dorsal side, 
thus reversing its relations. Barrande's orientation is here followed. 

Hyoliihes asteroideus, sp. nov. (PI. I. figs. 7 -9a.) 

Shell straight, gently tapering at about 1 in 5. Ventral face moderately convex, 
weakly angulated in centre and obscurely on each side near lateral edge, so that it shows 
two pairs of flattened surfaces, of which the median pair is the wider. Dorsal face 
weakly convex. Lateral angles subacute. Section plano-convex. Mouth oblique, with 
strongly arched, high dorsal lip. Ventral face with one faintly impressed longitudinal 
line along median keel, and traces of very delicate longitudinal striae over rest of 
surface, which is also crossed by more or less developed growth-ridges and striae. 
Dorsal face with strongly arched transverse growth-striae, most marked near mouth ; 
(longitudinal striae absent ?). 

Horizon and Locality. — Starfish Bed, Drummuck Group, Thraive Glen. 

Remarks. — Only imperfect shells of this species are known, but the formation of 
the ventral face is characteristic. One fragment with an apertural diameter of 9 mm. 
has a length of 20 mm., and the shell, when complete, must have measured about 45 mm. 
Eichwald's H. acutus* from the Orthoceras Limestone, has a somewhat similar shape 
and cross-section, but the ornamentation is different. t H. elegans, Barr.,| from 
Stage D, Bohemia, is certainly allied to our species. 

Hyolithes candidus, sp. nov. (PL I. figs. 10-13.) 

Shell straight, short, broad, tapering at about 1 in 2^. Dorsal face divided longi- 
tudinally into three areas, the lateral ones, which are each about one-quarter the width 
of the whole face, being flattened, gently inclined, and ornamented with 4-6 strong 
raised longitudinal lines ; the median area is of twice the width, slightly concave Or 
flattened, with very fine transverse arched striae and very delicate longitudinal striae, 
producing a minute cancellation of the surface. Dorsal lip of the mouth strongly 
arched. Ventral face rounded, gently convex, with rather distant fine longitudinal 
lines, about 20-25 in number, crossed by much more numerous transverse growth-lines. 
Lateral angles obtuse, rounded. 

Dimensions : — 

Length ........ 11-5 mm. 

Width of mouth . . . . . . . 4-5 „ 

Horizon and Locality. — Whitehouse Group, Whitehouse Bay. 

Remarks.— This species is rather imperfectly known, being based on the one 

* Eichwald, Leth. Ross., p. 1045, t. xl. figs. 13, 14. 

t Remel^, Zeitschr. deut.geol. Gesell., Bel. xli., p. 762, t. xxx. 

I Barrande, op. cit., p. 81, pi. xi. figs. 14-25, pi. xv. figs. 34-36. 



206 MR F. R. COWPER REED ON 

specimen showing the dorsal face and on several fragments. The dorsal face and shape 
of the shell resemble the Cambrian species H. signatidus, Novak, * from Stage C, but the 
longitudinal lineation is different. The ornamentation is more like that found in many 
species of Orthotheca in Bohemia, figured by Novak, but H. norvegicus, Kjerulf,f 
from Norway, may be compared with it. The minute cancellation is met with more 
generally in species of Orthotheca. The shape of the shell resembles the figure of a 
small unnamed species of Hyolithes from Tyrone, described by Portlock \ as the 
" internal bone of a Cephalopod." 

Hyolithes columbanus, sp. nov. (PI. I. figs. 14-1 Ad.) 

Shell large, gently curved ventralwards, tapering at about 1 in 4-5, with rather an 
obtuse blunt tip. Ventral face moderately convex, rounded, not angulated. Lateral 
angles acute, but not sharp. Dorsal face slightly convex. Mouth with dorsal lip 
gently arched, not high ; ventral lip straight. Cross-section unequally biconvex. Shell 
thick, with strong transverse growth-striae and ridges, especially on dorsal face, which 
also shows near lip about 20 impressed, widely distant, regular longitudinal striae, and 
a very delicate close longitudinal lineation over the whole surface ; on the ventral face 
the longitudinal striae are indistinct. 

Dimensions : — 

Length of fragment . . .55 mm. (estimated length of shell 80-90 mm.). 

Width of mouth . . 18 „ 

Horizon and Locality. — Drummuck Group, Thraive Glen. 

Remarks. — This species is not abundant. Its shape, curvature, and cross-section 
distinguish it from H. asteroideus, and its ornamentation from H. multi'punctatus (q.v.). 
Eichwald's H. lotus § is allied in curvature, cross-section, and apparently in ornamenta- 
tion, but tapers much more rapidly to the apex. 

Hyolithes girvanensis, sp. nov. (PI. I. figs. 15-17«.) 

Shell slender, straight, slowly tapering to sharp point at about 1 in 5 or 6. 
Ventral face angulated in middle, the two sides meeting at about 110°-120°, with 
narrow, flattened border at sides. Dorsal face flat, or nearly so. Cross-section triangular. 
Lateral angles acute, sharp. Ventral face with very fine longitudinal striation (rarely 
visible), and stronger transverse concentric growth-striae. Dorsal face with very slightly 
arched lip and strong transverse concentric growth-striae. 

Dimensions : — 

Length of shell . . . . . . . 17 - mm. 

Width of shell ....... 3-25 „ 

* Novak, "Rev. Palaeoz. Hyol. Bohm.," Abh. Icon. bohm. Gesell. Wiss., vii., Bd. iv. (Math.-Nat. Kl.), No. 
(1891), p. 23, t. iv. figs. 37-39. t Holm, op. cit., p. 87, t. iii. figs. 1-5. 

X PoBTLOOK, Geol. Hep. Londond., p. 470, pi. xxix.A. fig. 6. 
§ Eichwald, Leth. Ross., p. 1045, t. xl. fig. 1G. 



LOWER PALAEOZOIC HYOLITHID^ FROM GIRVAN. 207 

Horizon and Locality. — Balclatchie Group, Balelatchie. 

Remarks. — This species is allied to H. immemor (q.v.), from the same horizon, but 
differs by its more slowly tapering form, the non-chambered apex (so far as known), 
the more strongly angulated ventral face, and the sharp lateral angles. The longi- 
tudinal ornamentation of the ventral face is only very faintly seen in one specimen, and 
its existence is, indeed, open to doubt ; it is at any rate different to that of H. immemor. 

H. simplex, Salter,* is allied, but has not a sharply angulated ventral face, and 
tapers more rapidly ; the operculum referred to this species by Salter bears a con- 
siderable resemblance to that here mentioned under H. immemor, and the species 
occurs in the Lower Llandeilo of Shelve. H. vaginula, Salter, f is a shorter, broader 
shell, but has an angulated ventral face. 

H. triangularis (Portl.),| from Tyrone, is a larger shell, one of the type-specimens 
measuring just 40 mm. in length, but its rate of tapering and shape of cross-section 
seem identical ; the dorsal (flat) side is, however, stated to be longitudinally striated, 
while the ventral face has only transverse striae ; but all the specimens are very badly 
preserved, and the lineation, which is very unsatisfactorily seen, is at any rate coarser. 

Hijolithes immemor, sp. nov. (PI. I. figs. 18-21. ; PL II. fig. 1.) 

Shell straight, tapering at about 1 in 3|-, with the apical portion (i.e. the posterior 
fourth part of the shell) tapering more gradually and divided up by transverse 
partitions into 12-15 chambers. Dorsal face gently convex, with traces of narrow 
border. Dorsal lip gently arched. Lateral angles acute, but not sharp. Ventral face 
angulated, with strong submarginal raised longitudinal rib on each side, and two median 
weaker ribs along the central keel, but rest of surface smooth (imperfectly known). 
Surface of dorsal face marked by fine transverse growth-strise, with traces of 1-2 
marginal longitudinal lines occasionally visible. 

Dimensions : — 

Length . . . . . . . .21-0 mm. 

Width of mouth . . . . . . 5-0 „ 

Horizon and Locality. — Balclatchie Group, Ardmillan. 

Remarks. — This small shell has very constant proportions, but in no case is the 
ventral face well seen ; the marginal groove and border on the dorsal face are only 
visible when the shell is preserved, and then generally only in the apical half. The 
chambering of the apex and the slender, more tapering shape of this portion are 
remarkable, and have not been observed in other species from Girvan. The shell 
tapers less rapidly than H. ardmillanensis, but is probably closely allied to it, and 
imperfect casts are difficult to distinguish. Its relations to H. girvanensis have been 

* Salter, Mem. Geol. Surv., vol. iii., 1866, p. 352, pi. xi.B. figs. 22-26. 

t Ibid., p. 352, pi. x. fig. 14. 

I Portlock, Geol. Rep. Londond., p. 375, pi. xxviii.A. figs. 3 a-c. 



208 MR F. R. COWPER REED ON 

above mentioned. It is probable that Salter* refers to H. immemor in recording 
//. reversa (Salter) from the South of Scotland. In the latter species, originally 
described from North Wales,t there is no mention of a tapering chambered apex, but it 
seems closely allied in other respects. Eichwald (op. cit., p. 1045) believed that the 
Russian species, H. acutus, Eichw., which Holm J has fully described, from the 
Orthoceras Limestone, was a close ally of H. reversa. 

Amongst other Scandinavian species, it may be remarked that the ribbing of the 
ventral face of H. scalprum, Holm,§ is somewhat similar. 

Probably certain small opercula (PI. II. fig. 1) occurring in the same beds at 
Ardmillan are referable to this species. This type of operculum is subcircular in shape, 
conical, with an elevated excentrie apex situated close to the ventral margin ; the 
ventral face is triangular, short, flattened, more or less inclined to the general surface, 
and has a small median prominent fold below the apex, traversing it and extending 
to the margin. The general or dorsal surface of the operculum slopes down regularly 
from the apex and has one distinct regular concentric groove near the margin, and 
fine concentric strise above and below. In casts, two short, shallow, broad radial 
grooves, diverging from the apex at about 80° and widening outwards, traverse the 
dorsal portion of the shell, but do not extend more than two-thirds of the distance to 
the margin. These indicate muscular attachments, and correspond to those represented 
by Barrande || in opercula from Stage D in Bohemia. 

Hyolithes multipunctatus, sp. nov. (PI. II. figs. 2-26.) 

Shell straight, gently tapering at about 1 in 6-7. Dorsal and ventral faces gently 
convex, the ventral side rather more convex than the dorsal. Lateral edges sharp, acute. 
Mouth oblique ; dorsal lip slightly arched and swollen at margin. Surface of shell crossed 
by weak growth -ridges at irregular distances and by finer growth-lines and strise, and 
covered with small, round, shallow pits, situated at about two to three times their 
diameter apart. 

Dimensions : — 

Length (of broken fragment of shell) . . . 45 "0 mm. 

Width of mouth ....... 16-0 „ 

Horizon and Locality. — Balclatchie Group, Ardmillan. 

Remarks. — Only one broken specimen with the impression of its exterior is known, 
but this, luckily, has the mouth and ornamentation well preserved. The latter feature 
is peculiar, and does not appear to be of a secondary origin. It easily marks it off from 
other allied species, such as H. asteroideus. 

* Salter, Mem. Geol. Surv., vol. iii., 1866, p. 353, and woodcut 14, fig. 6, on p. 347. 

t Salter, in Murchison's Siluria, 2nd edit., p. 550, foss. 10, fig. 21. 

\ Holm, op. cit., p. 99, t. ii. figs. 29-36, t. vi. fig. 23. 

§ Holm, op. cit., p. 96, t. ii. figs. 46-50. 

|| Barrande, op. cit., pp. 95-97, pi. xiii. figs. a-e. 



LOWER PALAEOZOIC HYOLITHLD^E FROM GIRVAN. 209 



Hyolithes penkillensis, sp. nov. (PI. II. figs. 3-5.) 

Shell straight, short, broad, rapidly tapering at about 1 in 1\. Ventral face low, 
angulated (imperfectly known). Dorsal face flattened or very weakly convex, with 
narrow border distinctly marked off by groove, and with 1-3 strong, submarginal, 
longitudinal lines. Mouth with dorsal lip strongly arched. Lateral edges acute, sharp. 

Dimensions : — 

I. II. 

Length ... ... 15*0 mm. ll'O mm. 

Width at mouth . . .6-5,, 55 „ 

Horizon and Locality. — Penkill Group, Penkill. 

Remarks. — This broad, short form has much the shape of H. tardus, Barr., # from 
Stage G, but the margins of the dorsal face have the characters of H. norwegicus, 
Kjerulf.t H. obvius, Barr.,} from Stage E, tapers less rapidly to the point, and does not 
show the narrow borders on the dorsal face, as does the Girvan form. 

There is one specimen in the Hunterian Museum referable to this species from the 
same locality, but it tapers rather more slowly, measuring 17 mm. in length and 
6 '5 mm. in width at the mouth. 



Hyolithes subcrispatus, sp. nov. (PI. II. figs. 6-8.) 

Shell short, broad, straight, tapering rapidly at 1 in 2^. Dorsal face flattened, very 
weakly convex. Lateral angles sharp, acute. Ventral face unknown. Mouth with 
dorsal lip strongly arched. Surface of dorsal face crossed by strong, rounded, fairly 
regular and equal concentric growth-ridges or striae, and marked by 20-25 regular, 
straight, equal and equidistant raised thread-like lines, separated by broad interspaces. 
Occasionally two of the longitudinal lines seem more strongly developed, and divide the 
dorsal face into three equal longitudinal areas. 

Dimensions : — 

I. II. 

Length ...... 24 mm. 21 mm. 

Width at aperture . . . . 10 ,, 8 ,, 

Horizon and Locality. — Drummuck Group, Thraive Glen. 

Remarks. — This shell has more the shape and ornamentation of H. decipiens, Barr.§, 
but not such marked cancellation of the surface, nor a bent tip. H sulcatulus, Nov.,|| 
has a more slender shape and fewer longitudinal ribs. 

* Barrande, op. cit., p. 93, pi. vi. figs. 17-18 ; Novak, op. cit, p. 27, pi. iv. figs. 33-36. 

t Holm, op. cit., p. 87, t. iii. figs. 1-5. 

| Barrande, op. cit., p. 86, pi. xii. figs. 1-6. 

§ Barrande, op. cit., p. 80, pi. xii. figs. 33-37 ; Novak, op. cit., p. 16, t. iii. figs. 25-28. 

|| Novak, op. cit., p. 26, t. iii. fig. 22. 



210 MR F. R. COWPER REED ON 

Hyolithes sylvestris, sp. nov. (PI. II. figs. 9, 9a.) 

Shell short, straight, broad, tapering at about I in 2^. Ventral face unknown. 
Dorsal face weakly convex, with narrow raised border, separated by broad shallow 
groove ; surface ornamented with 30-40 fine, straight, thread-like longitudinal lines, of 
equal size and at equal distances apart, crossed by very delicate, closely placed, arched 
transverse strise. Dorsal lip strongly arched. 

Dimensions : — 

Length ........ 12'5 mm. 

Width at mouth . . . . . . .5-0,, 

Horizon and Locality. — Saugh Hill Group, Woodland Point. 

Remarks. — In shape this species, of which there is only one specimen, is indis- 
tinguishable from H. penkillensis, but the ornamentation of the dorsal face is sufficient 
to separate it. The shell is shorter than H crispatus (Boll. ),* but the ornamentation 
somewhat recalls it, though the Swedish form shows no raised border. 

Barrande's H. columnaris,i from Stages E and F, has a concave dorsal face, but a 
similar ornamentation. 

Perhaps Salter's Hyolithes anceps,\ from the Silurian of Eastnor Castle, is allied ; 
the shape and ornamentation appear to be similar, but the species is imperfectly known, 
and the type-specimen is indifferently preserved. 

Hyolithes, sp. ind. (a). 

An imperfectly known species of Hyolithes, somewhat resembling in shape 
H. immemor, is represented by four specimens from Penkill, in the Hunterian Museum. 
It is too poorly preserved to warrant giving it a specific name, but it is probably distinct 
and new. The shell tapers at about 1 in 4 ; the dorsal face is slightly convex and has a 
very slightly arched lip ; its surface is crossed by transverse strise ; there seems to be 
no border nor marginal longitudinal lines, as occasionally seen in H. immemor. The 
lateral angles appear to be sharp and acute. The ventral face is angulated or subangu- 
lated, and marked with a few longitudinal lines, apparently about 1 2 or rather more in 
number. 

One specimen measures 15 mm. in length, with an apertural width of about 4 mm., 
and another about 12*0 mm. in length, with an apertural width of about 3 25 mm. 

Hyolithes, sp. ind. (/3). 

Some imperfectly preserved examples of an indeterminable species of Hyolithes 
occur in the Mulloch Hill Group of Mulloch Hill. They are rather short, and taper 

* Holm, op. cit., p. 89, t. iii. figs. 36-52. 

t Barrande, op. cit., p. 78, pi. ix. fig. 13, pi. xii. fig3. 13-23. 

I Salter, Mem. Geol. Surv., vol. ii. pt. i., 1848, p. 355, pi. xiv. fig. 1. 



LOWER PALEOZOIC HYOLITHID^ FROM GIRVAN. 211 

at about 1 in 3-4. The dorsal face is flat, with a strongly arched mouth. No shell or 
ornamentation is preserved satisfactorily. One specimen measures 15 mm. in length 
and 4 mm. in width at the mouth. 

Hyolithes, sp. ind. (7). 

A small species of Hyolithes occurs in the Craighead Limestone, but it is too poorly 
preserved for specific determination. It is a broad, short form, tapering at about 1 in 3 ; 
the dorsal face seems to be flat, with a very slightly arched mouth, but no further 
features can be made out satisfactorily. The one specimen recognised measures about 
7 mm. in length and 2*5 mm. in width at the mouth. 

Subgenus Orthoiheca, Novak, 1887. 
Hyolithes (Orthotheca) subexcavatus, sp. nov. (PI. II. figs. 10-12.) 

Shell straight, slender, slowly tapering at about 1 in 8, with elongated, pointed 
apical portion showing chambers. Cross-section subtriangular. Ventral face high, 
sharply angulated, the sides meeting at nearly a right angle. Dorsal face concave, with 
swollen margins. Lateral angles acute, but rounded. Mouth straight, truncate. 
Ventral face ornamented with about 30 fine, straight, longitudinal thread-like lines, of 
equal size and equidistant, crossed by very numerous finer, closely placed stria?, giving 
the surface a cancellated appearance. Dorsal face with similar longitudinal lines and 
transverse striae. 

Dimensions : — 

Length . . . . . . . 18"5 mm. 

Width at mouth . . . . . . 2"5 „ 

Horizon and Locality. — Balclatchie Group, Ardmillan. 

Remarks. — In shape this shell much resembles H. girvanensis, but differs by 
tapering more slowly and by the absence of a dorsal lip and by the character of its 
ornamentation. In none of the specimens is the surface of the dorsal face well 
preserved, so that there is a little doubt as to the presence of the longitudinal 
lineation upon it. The chambering of the apical portion recalls H. immemor. In 
cross-section our shell resembles H. (0.) cor., Holm,* and H. (0.) excavatus, Holm,t 
of the Cambrian, but the ornamentation is somewhat like H. (0.) fiuctuosus, Holm,! 
from the Trinucleus Beds, and H. (0.) trinucleorum, Holm. § Our H. (0.) thraivensis 
tapers more rapidly, but otherwise is closely allied. 

There is a fairly good example of the upper portion of a shell from Desertcreat, 
Co. Tyrone, in the Jermyn Street Museum, which may be identified with this Girvan 
species. The shape and ornamentation are identical, and the specimen, when perfect, 
must have measured about 45 mm. long ; its width at the mouth is about 6 mm. 

* Holm, op. cit., p. 58, t. i. figs. 60-62. + Holm, op. cit, p. 61, t. v. figs. 67-71. 

I Holm, op. cit., p. 68, t. iii. figs. 53-57, t. v. figs. 88-91. § Holm, op. cit, p. 67, t. v. figs. 79-87. 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART II. (NO. 9). 31 



212 MR F. R. COWPER REED ON 

Hyolithes (Orthotheca) subornatus, sp. nov. (PL II. figs. 13-15.) 

Shell slender, straight, slowly tapering at about 1 in 8, triangular in cross-section. 
Ventral face strongly angulated at nearly a right angle, high, with strong median 
longitudinal rib and 6-8 fine, straight, thread-like longitudinal lines on each side (with 
2-3 very delicate ones between adjoining ones), and a strong marginal or supra-marginal 
longitudinal rib near the lateral edge. Dorsal face apparently flat or slightly concave 
and crossed by very fine transverse striae. Mouth straight. 

Horizon and Locality. — Whitehouse Group, Whitehouse Bay (and Shalloch Mill?). 

Remarks. — The peculiar ornamentation in this species distinguishes it from all 
others in the Bala Beds. One specimen from Whitehouse Bay measures about 30 mm. 
in length, but the tip is missing, so that its original length must have been about 35 
mm., and its diameter at the mouth is about 4 mm. 

Some poorly preserved specimens from Shalloch Mill probably belong to this species, 
but seem to taper rather more rapidly. 

Hyolithes (Orthotheca) thraivensis, sp. nov. (PI. II. figs. 16-23a.) 

Shell very slightly and regularly curved dorsalwards, or straight, gently tapering at 
about 1 in 6 or 7, markedly triangular in cross-section, nearly as high as wide. Ventral 
face strongly convex, high, more or less angulated. Dorsal face flat or slightly concave. 
Lateral angles rounded, obtuse. Mouth straight, not oblique ; dorsal lip straight, not 
arched. Surface of ventral face marked with 30-36 longitudinal equal or subequal 
raised lines, separated by broader interspaces, and crossed by subequal, more closely 
placed and more numerous but weaker raised lines, forming granules at the intersections, 
and giving a cancellated appearance to the surface ; the 1 or 2 longitudinal lines on the 
rounded lateral angles are rather stronger than the rest. Dorsal face crossed by straight 
transverse lines similar to and continuous with those on the ventral face, but without 
longitudinal lines. 

Dimensions : — 

I. II. 

Length about ..... 33*0 mm. 37 -0 mm. 

Width at mouth about .... 5*0 „ 6"0 ,, 

Horizon and Locality. — Drummuck Group (? including Starfish Bed), Thraive Glen. 

Remarks. — This beautifully ornamented species seems abundant in the Drummuck 
Beds of Thraive Glen, and is usually well preserved and has the shell still attached. 
//. (Orthotheca) trinucleorum, Holm,* maybe compared with this species, and it comes 
from a homotaxial horizon. The ornamentation in H. (O.) thraivensis is coarser than 
in //. (O.) snbexcavatus, and the shell tapers more rapidly, and is usually somewhat 
curved. 

* Holm, op. cit., p. 67, t. v. figs. 79-87. 



LOWER PALEOZOIC HYOLITHID^E FROM GIRVAN. 213 

Hyolithes (Orthotheca) thraivensis, var. granuliferus. (PL II. figs. 24, 25.) 

Shell straight, triangular, slowly tapering at about 1 in 6. Ventral face high, 
angulated, its sides meeting at about a right angle. Dorsal face slightly concave. 
Lateral angles obtuse, rounded. Mouth straight. Ventral face bearing one strong 
median raised longitudinal line along ridge and 6-8 finer raised lines on each side, each 
line bearing a single row of small equidistant tubercles. The longitudinal lines are not 
all of equal strength or at equal distances ; some are finer than others and are not 
continued to the point, only extending about half or two-thirds the length of the shell ; 
the interspaces between the lines are wide, and very delicate transverse lines cross them. 
The rounded lateral angles bear several similar longitudinal lines, but the dorsal face 
seems devoid of them, and only shows very faint transverse strise. 

Horizon and Locality. — Drummuck Group, Thraive Glen. 

Remarks. — The one internal cast and external impression on which this variety is 
based differ from the typical form by the granules on the longitudinal lines being 
fewer and so much larger as to be w T orthy of the designation of tubercles. The longi- 
tudinal lines also are fewer and less regular, and the concentric strise much finer and 
fewer. The fragment measures about 1175 mm. in length, with an apertural diameter 
of about 4 '25 mm. 

Hyolithes (Orthotheca?), sp. ind. 

An elongated, narrow, slowly tapering species from Penkill is represented by one 
poor and imperfect specimen in the Hunterian Museum, Glasgow. The fragment 
measures 32 mm. in length, but the tip is broken off; at its broader end it measures 
3 "25 mm. in width ; its true rate of tapering is about 1 in 11. The impression of the 
one face preserved shows that this face ( ? the dorsal) must have been gently convex, 
and there are traces of faint transverse, scarcely arched striae running across it. No 
other features can be detected. 

Genus Ceratotheca, Novak, 1891. 

Ceratotheca ? subuncata, sp. nov. (PI. II. figs. 26-28.) 

Shell cornuate, strongly curved back at about half its length, and thence rather 
rapidly tapering to point, the curvature being such that the posterior part of the shell 
is bent back in a regular, nearly semicircular curve, so that the apical extremity lies at 
about two-thirds to three-fourths the length of the shell. Section of shell subcircular 
or elliptical. Mouth at right angles to shell, not oblique ; margin of mouth slightly 
sinuous. Shell thin ; surface ornamented with very fine concentric growth-striae 
and with very delicate closely placed longitudinal striae (observable near mouth), which 
on crossing the transverse strise are raised into minute granules. 



214 



MR F. R. COWPER REED ON 



I. 


II. 


Til. 


15-0 


22-0 


21-0 mm 


8-5 


10-5 


1 1 -0 mm 



Dimensions : — 

Length of shell (to bend) . 
Diameter of mouth 

Horizon and Locality. — Penkill Group, Penkill. 

Remarks. — The specimens are mostly rather crushed and imperfect, and the surface 
ornamentation is rarely visible. The species which most resembles it is Ceratotheca 
unguiformis, Novak, # from Stage Ee 1 in Bohemia, but the rate of tapering in the latter 
is less rapid and the curvature of the shell begins nearer the apex ; the ornamentation 
also is different. 

The inclusion of C. subuncata and the following species, C. balclatchiensis, in 
the genus Ceratotheca, appears to me rather doubtful. However, it is impossible to 
establish a new genus on such indifferent material, and it seems best to leave them 
associated in it for the present. The resemblance of C. subuncata in general shape to 
some species of Eccyliomphalus may be noted, and even Lindstrom's Euomphalus 
gotlandicus t has a certain degree of similarity. 



II. 


III. 


25-0 


18-0 mm. 


9-0 


7 - mm. 



Ceratotheca ? balclatchiensis, sp. nov. (PI. III. figs. 1-3.) 

Shell conical, slightly curved ventralwards, rather rapidly tapering at about 1 in 3. 
Cross-section subcircular to subelliptical. Ventral face strongly convex, rounded. 
Lateral angles rounded, obsolete and indefinite. Dorsal face rather less convex than 
ventral. Mouth straight or with margins slightly sinuous, at right angles to shell ; 
no arched dorsal lip. Shell lustrous, corneous, thin, smooth, or with irregular faint 
concentric growth-ridges and striae. 

Dimensions : — 

I. 
Length ..... 21*5 
Diameter at mouth . . . 8"0 

Horizon and Locality. — Balclatchie Group, Balclatchie. 

Remarks. — None of the specimens are very well preserved or complete, but it is an 
abundant form at Balclatchie, and the numerous examples enable one to give a fairly 
complete diagnosis. The affinities of this species are doubtful. It is much less com- 
pressed dorso-ventrally than Hyolithes aduncus, Barr., from Stage E ; the lateral angles 
are less evident and the shell tapers less rapidly. The thin, black, corneous nature of 
the shell makes it doubtful if this organism is rightly referred to the Pteropoda at all, 
but it hardly seems possible to regard it as annelidan in nature, and it may provision- 
ally be referred to Ceratotheca. From its shape it looks as if it might be more correctly 
placed with the Scaphopoda, but the nature of the shell is against this reference. 

* Novak, op. cit., p. 33, t. i. figs. 23, 24. 

t Lindktrum, Silur. Gastrop. Pterop. Gotland (1884), p. 139, pi. xiii. figs. 19-31. 



LOWER PALEOZOIC HYOLITHID^ FROM GIRVAN. 215 

Genus Plerotheca, Salter, 1852. 

The genus Pterotheca, which was established by Salter # to contain an Irish fossil 
described by Portlock as Atrypa transversa, is represented not only by British 
species, but also by others from America and Bohemia.t Its true position and affinities 
have been a matter of uncertainty : Zittel, as recently as 1903, retained it in the 
Hyolithidce, but it is believed by Ulrich and Scofield \ to belong to their family 
Carinaropsidod, which is defined by them as follows ; — " Symmetrical, almost patelliform 
shells, the aperture being greatly expanded : apex small, involute, overhanging the 
posterior margin, consisting of no more than two volutions. Within the aperture a 
broad concave septum. Anterior lip with a central emargination." The type of the 
genus Carinaropsis § is C. carinata, Hall, of the Trenton Group, and the close affinities 
of Carinaropsis with Bellerop>hon and Pterotheca are clearly pointed out by Ulrich 
and Scofield. 

Pterotheca agrees with Carinaropsis in a striking manner, and possesses the 
characteristic internal septum, the broadly expanded mouth, and the carinated back, but 
differs by the shell being simply arcuate with an incurved beak, instead of being coiled 
into two volutions, and also by possessing a pair of small vertical septa in the apical 
extremity, dividing the cavity into three portions. This peculiar septate apical region 
is considered by the same authors to be of generic importance only, but sufficient to 
make it doubtful whether Pterotheca is a Pteropod, and should not rather be classed 
with Carinaropsis as a family of the Docoglossa. Novak [[ considered that his genus 
Pterygotheca somewhat recalled Salter's Pterotheca, but it is difficult to see any 
structural affinity, or even to recognise much superficial resemblance. 

Salter's definition of the genus Pterotheca is short : — " Shell transversely oval, 
bilobed, with wavy sides and a strong median keel. Ventral plate short, narrow, flat." 
Only a few species have been described, and the type is Pt. transversa (Portl.).H 
The genus is stated to occur in Bohemia and North America, and it ranges through 
the Ordovician and Silurian. The best-known British species is Pt. corrugata, Salter, 
1852, which is recorded from the "Caradoc" rocks of Dolbenmaen, near Criccieth.* # 
Pt. undulata, Salter,"!"! from Carnarvonshire, has apparently never been figured or 
described. Pt. avirostris, Salter (MS.), is from the Woolhope and Upper. Llandovery 
Beds. 

* Salter, Brit. Assoc. Reports, 1852, Trans. Sect., p. 61. 

t Barrande, op. cit., p. 99, pi. xv. figs. 1-6. 

I Ulrich and Scofield, "Lower Silur. Gastr. Minnesota" (Geol. Nat. Hist. Surv. Minnesota, vol. iii. pt. ii., Final 
Rep., " Palseont," 1897, chap, x.), pp. 857, 926. 

§ Hall, Palaont. New York, vol. i., 1847, p. 183. 

|| Novak, "Rev. Palseoz. Hyolith. Bohmens," Abhandl. ton. bbhm. Gese.ll. Wissen., vii., Bd. iv. (Matli.-nat. Kl.), 
No. 6, 1891, p. 45, t. iii. figs. 36-47. 

1" Portlock, Geol. Rep. Londond., p. 455 (Atrypa transversa) ; Salter, Brit. Assoc. Rept., 1852, Trans. Sect., p. 61 ; 
Murchison's Siluria, 5th ed., 1872, foss. 41, fig. 4, p. 199. 

** Salter, Mem. Geol Surv., vol. iii., 1866, p. 353, fig. 18. 

tt Salter, op. cit, 1866, p. 274. 



216 MR F. R. COWPER REED ON 

Pterotheca drummuckensis, sp. nov. (PI. III. figs. 4, 5.) 

Shell subcircular to transversely subcordiform, rather broader than long. Apertural 
margin emarginate, with broad, rounded lateral lobes, and broad, angular median notch 
hot ween them. Median fold narrow, sharp, acutely angular, not projecting on apertural 
margin, but corresponding with angular notch. Lateral folds very low, wide and inde- 
finite, passing into flattened lateral portions of shell. Posterior margins of shell meeting 
at apex at about 100°-110°. Pair of narrow shallow grooves diverging forwards from 
apex at about 45°, halfway between median ridge and lateral margins of shell. 

Surface marked with regular strong corrugations, concentric to apertural margin, and 
bending back sharply in median notch so as to form re-entrant angle of about 50° on 
ridge of keel. 

Dimensions : — 

I. II. 

Length ...... 20-0 150 mm. 

Width ...... 20-0 18-0 „ 

Horizon and Locality. — Drummuck Group, Thraive Glen. 

Remarks. — This shell, though only known by imperfect specimens, differs from the 
next described species, Pt. girvanensis, by its shape, its non-projecting median keel, the 
angle of divergence of the grooves on the surface, the acutely emarginate apertural 
margin, and the regular corrugation. 

Pterotheca girvanensis, sp. nov. (PI. III. figs. 6-13.) 

Shell transversely subelliptical, widest towards the front. Apertural margin sinuous, 
projecting in middle in short subangular tongue, excavated slightly on each side, and 
with rounded, more or less projecting lateral lobes, forming the angles and merging into 
the lateral margins, which are rounded. Apex small, sharp, incurved. Surface of shell 
gently convex, elevated into sharp median longitudinal fold, proceeding from apex with 
increasing height, width, and acuteness to the apertural margin, where it stands up 
prominently and projects as the median tongue ; a pair of broader but low, rounded, 
indefinite lateral folds run forwards from the apex divergently, almost dying out on the 
apertural margin, where they form the lateral lobes. Two shallow definite grooves, 
slightly curving outwards, diverge from the apex at about 60°-70°, and run forwards 
towards the outer lateral angles of the apertural margin, nearly or quite reaching it, and 
they lie on the outer slopes of the lateral folds. These grooves correspond internally 
with low ridges or septa. Internal concave plate present, extending about two-thirds 
the length of the shell. Surface of shell minutely punctate, and with more or less de- 
veloped concentric corrugations and strise following the outline of the apertural margin. 

Dimensions : — 

I. IT. III. 

Length ..... 15-0 90 13"0 mm 

Width ..... 20-0 12-5 20"0 „ 



LOWER PALEOZOIC HYOLITHID^ FROM GIRVAN. 217 

Horizon and Localities. — Balclatchie Group, Balclatchie and Ardmillan. 

Remarks. — This species is undoubtedly closely allied to Pt. corrugata, Salter. 
The median slit which the latter species is represented as possessing near the anterior 
margin is apparently due merely to the breaking off of the acute carina, as precisely 
the same appearance is met with in damaged specimens of this Girvan form. There are 
three lateral folds represented by Salter on each side of the central one, and the con- 
centric strise are said to meet the keel abruptly at a right angle, whereas in ours they 
pass over it continuously, following the outline of the apertural margin. 



Pterotheca mullochensis, sp. nov. (PL III. figs. 18-20.) 

Shell transversly elliptical to subcircular. Apertural margin arched forwards, not 
emarginate, with small projecting triangular tongue in middle. Surface gently convex. 
Apex small, incurved. Median fold angulated, acute, narrow and weak near beak, but 
increasing in height and width to front margin, where it projects slightly as tongue. 
Lateral lobes on apertural margin and lateral folds on surface obsolete, but halfway 
between median fold and lateral margin of shell the surface is angulated, being bent 
down rather suddenly on each side, along lines diverging from the apex at about 60° ; 
the angulations die out towards the anterior margin, curving gently outwards, or 
running nearly straight forwards ; they correspond near their origin to the course of the 
thin internal divergent septa which start from the apex and extend about one-third to 
one-half the length of the shell. Internal plate gently concave, extending forwards 
rather more than half the length of shell. Surface marked with numerous moderately 
strong concentric corrugations and stria?. 

Dimensions : — 

I. II. III. 

Length ..... 24-0 22-5 23-0 mm. 

Width ..... 26-0 27-0 c. 25-0 „ 

Horizon and Locality. — Mulloch Hill Group, Mulloch Hill ; and (?) Saugh Hill 
Group, Newlands. 

Remarks. — The shape of the anterior margin, the obsolescence of the lateral lobes 
and folds, the angulation of the lateral surfaces and general shape of the shell, dis- 
tinguish it from Pt. girvanensis and Pt. drummuckensis. Portlock's Pt. transversa 
from Desertcreat, the type of which is in the Jermyn Street Museum, much resembles 
Pt. mullochensis in surface characters, but is broader and more transversely elliptical in 
shape, its length to its width being as 2-3 ; and it possesses, near the beak, a small 
additional pair of very short septa between the two long divergent ones, and the latter 
bend away at rather a wider angle, though this may be due to crushing. The con- 
centric rugse are less developed, the concentric markings being rather of the nature of 
fine wrinkles and striae. 



I. 


II. 


9-5 


15*5 mm. 


11-5 


21-0 mm. 



218 MR F. R. COWPER REED ON 

Pterotheca multidecorata, sp. nov. (PI. III. figs. 14-16.) 

Shell semi-elliptical, with long, strongly arched apertural margin and short postero- 
lateral margins meeting at apex at wide angle (about 130°-150°); apertural margin 
narrowly and not deeply, but acutely, emarginate in centre. Surface of shell gently 
convex ; median fold high, angulated, acute, with narrow raised keel, increasing 
gradually in width and somewhat in height to anterior margin, where it ends in the 
median notch ; lateral folds low, rounded, curving slightly outwards and projecting as 
rounded lobes on margin, well marked off from median ridge and from lateral portions 
of shell, which form broader and lower outer rounded folds ; the grooves separating 
these outer from the inner lateral folds diverge at rather less than 45°. Surface of 
shell marked (l) by a few concentric corrugations and striae bending back acutely at a 
re-entrant angle of 45°-50° on median fold, and (2) by fine radial striae, slightly wavy, 
and increasing in number towards the margin, containing along them a row of closely 
set minute punctae. Internal septa indistinct, thin, gently curved, less than the length 
of shell, diverging from apex at about 30°-45°, and running in outer lateral grooves. 

Dimensions :■ — 

Length ...... 

Width ...... 

Horizon and Locality. — Penkill Group, Penkill. 

Remarks. — The ornamentation of this shell easily marks it off from all the other 
species, somewhat resembling that in the brachiopod Porambonites. The two pairs 
of lateral folds and shape of the shell are additional distinguishing features. Pt. 
avirostris, Salter (MS.), from the Upper Llandovery of Bog Mine, Shelve, and Damery 
Bridge, Tortworth, is only known as internal casts. It is almost circular in outline, and 
does not show clearly two pairs of lateral folds, but the high angulated narrow keel and 
the small angle of divergence between the internal septa constitute points of resemblance 
between it and Pt. multidecorata which are worthy of notice. 

Pterotheca simplex, sp. nov. (PI. III. figs. 17, 17a.) 

Shell small, transversely semi-elliptical, with strongly arched apertural margin. 
Surface gently convex, with one broad, slightly elevated subangular median fold, 
increasing rapidly in width from the beak to the apertural margin, where it projects 
slightly and has a width equal to nearly half the shell ; it is bounded laterally by 
weakly impressed shallow grooves, curving gently outwards to margin, and diverging 
from apex at about 45°. Lateral folds obsolete. Concentric corrugations and striae few 
and faint, strongest near apex. 

Dimensions : — 

Length . . . . . . . . 5 25 mm. 

Width . . . . 7-0 „ 



LOWER PALAEOZOIC HYOLITHID^ FROM GIRVAN. 



219 



Horizon and Locality. — Balclatcliie Group, Balclatchie. 

Remarks. — The simplicity of the lobation of this small shell distinguishes it from 
Pt. girvanensis, and it does not seem possible to regard it as merely the young of that 
species. It more resembles the Bohemian species, Pt. bohemica, Barr.* The 
American species, Pt. expansa (Emmons),t may also be compared. 



List of Species. 



Hyolithes ardmillanensis, sp. nov. 

,, asteroideus, sp. nov. 

„ candidus, sp. nov. 

., columbanus, sp. nov. 

„ girvanensis, sp. nov. 

,, immemor, sp. nov, 

„ multipunctatus, sp. nov. 

,, penkillensis, sp. nov. 

,, subcrispatus, sp. nov. 

,, sylvestris, sp. nov. 

,, sp. ind. (a). 

., sp. ind. (f3). 

„ sp. ind. (y). 



Hyolithes (Orthotheca) subexcavatus, sp. nov. 

,, ,, subornatus, sp. nov. 

„ ,, thraivensis, sp. nov. 

,, ,, „ var. granuliferus. 

1 sp. ind. 
Ceratotheca 1 subuncata, sp. nov. 

„ 1 balclatchiensis, sp. nov. 

Pterotheca drummuckensis, sp. nov. 

,, girvanensis, sp. nov. 

,, mullochensis, sp. nov. 

,, multidecorata, sp. nov. 

„ simplex, sp. nov. 



Craighead Limestone — 
Hyolithes, sp. ind. (y). 

Balclatchie Group — 

Hyolithes ardmillanensis. 

„ girvanensis. 

„ immemor. 

„ multipunctatus. 

„ (Orthotheca) subexcavatus. 
Ceratotheca 1 balclatchiensis. 
Pterotheca girvanensis. 

„ simplex. 

Whitehouse Group — 

Hyolithes candidus. 

,, (Orthotheca) subornatus. 

Drummuck Group — 

Hyolithes asteroideus. 
„ columbanus. 

* Barrande, op. cit., p. 104, pi. xv. figs. 1-6. . 
t Emmons, Geol. Rep. 2nd Distr. New York, 
Pakeont. iii. (1902), p. 179, pi. xii. fig. 35. 

TRANS. ROY. SOC. EDIN., VOL. XLVII. 



Stratigraphical Distribution of the Species. 

Drummuck Group — continued. 



var. granuliferus. 



Hyolithes subcrispatus. 

„ (Orthotheca) thraivensis. 

>; >) ii 

Pterotheca drummuckensis. 

Mulloch Hill Group — 

Hyolithes, sp. ind. (/3). 
Pterotheca mullochensis. 

Saugh Hill Group — 

Hyolithes sylvestris. 

,, sp. ind. (a). 
? Pterotheca mullochensis. 

Penkill Group — 

Hyolithes penkillensis. 

„ (Orthotheca 1) sp. ind. 
Ceratotheca 1 subuncata. 
Pterotheca multidecorata. 



1842, p. 397, figs. 109-112 ; Weller, Geol. Sun. New Jersey, 
PART II. (NO. 9). 32 



220 MR F. R. COWPER REED ON 

Remarks on the Stratigraphical Distribution. 

It is noticeable that the greatest variety of species is met with in the Balclatchie 
Group. In the underlying Stinchar Group only one species, represented by a single 
individual, is so far known. The Whitehouse Group has yielded no more than two 
species, while the Drummuck Group is particularly rich in members of the family. 

In the Silurian Beds it is remarkable that the Penkill Group contains a greater 
abundance than the earlier Silurian horizons. 

This irregular development of specific types does not appear to be directly related 
to the lithological characters of the strata, for though the argillaceous Balclatchie rocks 
have the maximum of species, the sandy Drummuck Beds have many more types than 
the fine-grained shaly Whitehouse Group. 

On the other hand, the arenaceous Mulloch Hill and Saugh Hill Groups are very 
poor in representatives of the family, while the Penkill mudstones contain a compara- 
tive abundance of forms. 

Other factors than the coarseness of the sediments must have been at work in 
determining the relative abundance of species. It cannot be alleged that their absence 
or rarity is due to their non -preservation in a coarse matrix, for the evidence of the 
Drummuck Beds proves that sandy deposits lend themselves just as well as clays to 
their conservation. Their independence of the rock-composition is as noteworthy in 
Bohemia as in the Girvan district. 

In addition to the relative number of species on the various horizons, we must have 
regard also to the relative number of individuals of each species, as evidenced by the 
specimens in the collections. In this case, also, the maximum is found in the 
Balclatchie and Drummuck Groups. On the former horizon, H. ardmillanensis, 
H. girvanensis, and H. immemor are represented by so many specimens that they easily 
outnumber all the other forms. In the Drummuck Beds the commonest species are 
H. (Ortho.) ihraivensis and H. asteroideus, but the former is by far the most 
abundant. 

With respect to the Silurian species, there are very few examples of any of them, 
and less than a dozen have come under my notice from the Mulloch Hill and Saugh 
Hill Groups. In the Penkill Group individuals are less rare, but pteropods do not 
seem to be at all abundant members of the fauna. 

We are generally precluded from making any estimate, even with approximate 
accuracy, of the actual relative abundance of species and individuals of any zoological 
group of fossils in the beds deposited at a given period, the obstacles being the paucity 
of material and the unsystematic method of collecting, or accidental manner by which 
most museums have obtained their fossils. But in this case Mrs Gray's regular visits and 
careful collecting from the sjpne district and horizons for many years past justify us in 
considering the above conclusions as to the proportional distribution of the Hyolithidae 
in the beds of the Girvan area to be a closer approximation to the truth than usual. 



LOWER PALAEOZOIC HYOLITHLD./E FROM GIRVAN. 221 

Relations of the Girvan Hyolithid^e to those of other Areas. 
It is a remarkable fact that all the Girvan species, except one or possibly two which 
occur in Co. Tyrone, appear to be peculiar and new. Afber a careful comparison 
with all the well-known English and Welsh species, it does not seem possible to regard 
any as identical, and the relationship of any of them is not very close. Many, however, 
of the Girvan forms are allied to Scandinavian ones, of which we have such a full and 
detailed knowledge owing to Holm's memoir. A few show affinities with Bohemian 
species. It may be noticed that amongst the Girvan Conularice, Miss Slater * 
recognised one species identical with a Scandinavian form, and more than one possessing 
considerable affinities with those from that region. 

Conclusion. 

By the foregoing study of the Hyolithidce the Girvan area is seen to contain an 
unusually rich and varied assemblage of members of the group. The interest of 
them is heightened by the local restriction and peculiarity of the species, and by the 
frequently good state of preservation of the material, which renders it possible to define 
their characters with greater precision than is usual in dealing with members of the 
same group from other British localities. 



EXPLANATION OF PLATES. 
Plate I. 



Figs. 1-6. Hyolithes ardmillanensis. 1, ventral face ; 2, dorsal face ; 2a, cross-section ; 3, dorsal face ; 

4, ventral face; 5, dorsal face; 6, operculum. (All x 2.) 
7-9a. Hyolithes asteroideus. 7, ventral face; 7a, cross-section; 8, dorsal face; 9, ventral face; 

9a, cross-section. (All x 2.) 
10-13. Hyolithes candidus. 10, dorsal face; 11, 12, ventral face (all x 2h) ; 13, ornamentation of 

ventral face ( x 5). 
14-14'i. Hyolithes columbanus. 14, dorsal face; 14a, side view ; 146, ventral face; 14c, cross-section 

(all x 2) ; 14a", portion of dorsal face near lip ( x 3). 
15-17«. Hyolithes girvanensis. 15, ventral face; 16, dorsal face; 17, 17a, dorsal and ventral faces of 

same shell. (All x 2£.) 
18-21. Hyolithes immemor. 18, 18a, dorsal face, impression and cast; 19, dorsal face; 20, ventral 

face (three-quarter view, right side); 21, dorsal face. (All x 2|.) 

Plate II. 

Fig. 1 . Hyolithes immemor. Operculum ( x 2^). 

Figs. 2-26. Hyolithes multipunctatus. 2, ventral face ; 2a, cross-section (both x 2) ; 2b, ornamentation 
( x 3). a 

„ 3-5. Hyolithes penkillensis. Dorsal faces ( x 2|). 

* Slater, Mon. Brit. Conularice (Palseont. Soc), 1907, pp. 16, 21, 31. 



I - 


c.,4 


-2- - 


Cr. ^9 




Cr-U^ 


1 • 


Q *>~7l 


fr 


It V-h 


s -. 


q.*t~> 1 


\o ~- 


^¥>^ 


n = 


C i^c 


/-z. 


«c^ 



222 LOWER PALEOZOIC HYOLITHIDJE FROM GIRVAN. 

Figs. 6-8. Hyolithes subcrispatus. Dorsal faces ( x 2). 
,, 9-9a. Hyolithes sylvestris. 9, dorsal face ( x 2|) ; 9a, portion of surface ( x 5). 
,, 10-12. Hyolithes (Orthotheca) subexcavatus. 10, dorsal face; 10a, cross-section; 11, side view of 

ventral face (all x 2|) ; 12, portion of surface of ventral face ( x 5). 
,, 13-15. Hyolithes (Orthotheca) subornatus. 13, ventral face ( x 2£) ; 13a, portion of surface of ditto 

( x 5) ; 14, interior from dorsal side ( x 2£) ; 15, ditto ( x 2). 
,, 16-23a. Hyolithes (Orthotheca) thraivensis. 16, side view; 17, dorsal face; 18, ditto; 19, ventral 

face; 20, side view of ventral face; 21, cross-section; 22, ventral face; 23, ditto; 

23a, cross-section. (All x 2.) 
,, 24, 25. Hyolithes (Orthotheca) thraivensis, var. granuliferus. 24, impression of ventral face ( x 2£) ; 

24a, portion of surface of ditto ( x 5) ; 25, cast of same specimen ( x 2|). 
,, 26-28. Ceratotheca 1 snbuncata. Side views ( x 2). 



Plate III. 

Figs. 1-3. Ceratotheca 1 balclatchiensis. 1, side view ; 2, ventral view ; 3, ditto, showing apex. 

(Allx 2.) 
,, 4-5. Pterotheca drummuckensis. 4, cast ; 5, impression showing internal plate. (Both x 2.) 
,, 6-13. Pterotheca girvanensis. 6, showing internal plate ; 10, showing marginal projection of carina ; 

11, internal cast; 12, impression of part of exterior; 13, showing internal plate (all x 2). 

6-10 from Balclatchie ; 11-13 from Ardmillan. 
,, 14-16. Pterotheca multidecorata. 14, 15, 16 (all x 2); 15a, portion of surface ( x 5). 
,, 17,17a. Pterotheca simplex. 17a, cross-section ; 17, 17a (all x 2|). 
,, 18-20. Pterotheca muttochensis. (All x 2.) 



ans .Roy. Soc.Edin . 

MR F.RCOVVPER REED On"lOWER PALAEOZOIC HYO L1TH 1 DJE FROM Gl RVAn" PLATE 1 



Vol. XLVIL 






1 


■ -a 




: '\m 




i f 

f. 





14 a. 



14Hb 



21 



J.Green del.lith.et imp. 



^ans . Roy. Soc .Edm . Vol . XLVII . 

MR F. R.COWPER REED On"lOWER PALAEOZOIC HYOL1TH1D/E FROM GIRVAN" PLATE II . 




J 






1 


■IS 

fi i 

m i 


4 m 




v).<JFf| 


t * 





10. 


L 




■ "1 


1 ' ."■ 


B i 
H i 


1 


■Pi ' 

■If i 





<J. Green del.lith.et imp. 



.ns Rx>y. Soc .Edin. Vol . XLV1I. 

M«F.R.COWPER REED On"lOWER PALAEOZOIC HY0L1TH1D>E FROM G1RVAN" PLATE 111. 




20. 



J.Green. del lith.et imp.' 



( 223 ) 



X. — Report upon the Anatomy and Embryology of the Penguins collected by 
the Scottish National Antarctic Expedition, comprising : (1) Some Features 
in the Anatomy of the Penguin ; (2) The Embryology of the Penguin : A 
Study in Embryonic Regression and Progression. By D. Waterston, M.A., 
M.D., F.R.C.S.E., and A. Campbell Geddes, M.D. (From the Laboratory of 
the Anatomical Department, The University, Edinburgh.) (With Three Plates.) 

(MS. received March 11, 1909. Read February 3, 1908. Issued separately October 21, 1909.) 

Among the specimens collected by the Scottish National Antarctic Expedition and 
brought back to this country was a series of the embryos of penguins of various species, 
collected at different stages of development by Mr R. N. Rudmose Brown and Dr J. H. 
Harvey Pirie. These embryos were, for the greater part, of the species Gentoo 
(Pygoscelis papua, Forst.), the remainder of the species Adelia (Pygoscelis adelice, 
Hombr. and Jacq.). Two Emperor penguins, which had been preserved for examination 
by the injection of the blood-vessels, were also brought home. Through the kind- 
ness of Dr W. S. Bruce these specimens were entrusted to one of us (D. W.) for 
examination and report, but as the work involved in this proved to be very 
extensive, Dr A. Campbell Geddes was asked to undertake a share in it, and this 
he agreed to do. 

This report contains the results of the combined investigations. 

PART I. 

ON THE ANATOMY OF THE ADULT PENGUIN. 
Introductory. 

Upon the return of the Challenger Expedition to this country, the late Professor 
Morrison Watson had handed over to him the valuable collection of adult penguins 
which had been made. Upon these he based his classical memoir, " A Report on the 
Anatomy of the Spheniscidse collected during the Voyage of H.M.S. Challenger" 
{Challenger Reports, vol. vii. p. 1). To that report little, if anything, can be added. 
There are, however, two points which a detailed dissection of the adult penguin has 
brought to light, which seem to us to be worthy of being placed on record. These are, 
first, the peculiarities of the cervical portion of the vertebral column ; second, some points 
in connection with the arrangement of the fascial layers- in relation to the pectoral 
muscles. In all the other points illustrated by our dissections we can merely corrobo- 
rate the description given by Professor Watson. 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART II. (NO. 10). 33 



224 DR D. WATERSTON AND DR A. CAMPBELL GEDDES ON EMBRYO PENGUINS 

Section 1. — Some Peculiarities of the Cervical Portion of the 

Vertebral Column. 

The cervical vertebrae are thirteen in number : the portion of the vertebral column 
which they make up is remarkable for the development of its antero-posterior curves. 
In the lower half there is an extraordinary antero-posterior curve the convexity of 
which is directed forward ; in the upper half there is a second antero-posterior curve the 
convexity of which is directed backward. 

These curves are present in all birds ; in none, however, do they approach in intensity 
the curvatures found in the penguin. The lower curvature is so pronounced that the 
bodies of the vertebrae come in contact with and fill up the angle formed by the limbs 
of the clavicle, and, in the living bird, actually cause a projection on the front of the 
neck. As a result the trachea and oesophagus, instead of passing into the thorax in 
front of the vertebral column, are pushed away to the right side and actually lie on a 
plane posterior to the vertebrae of the convexity. The knowledge of this extraordinary 
condition is not new ; it is fully described by Professor Watson, but the point seems to 
have been entirely overlooked, for in all the specimens in museums that we have had 
the opportunity of seeing the cervical vertebrae are incorrectly mounted. 

Professor Morrison AVatson was of opinion that these curvatures were associated 
with the maintenance by these birds when on land of the erect attitude, and that they 
served to bring the centre of gravity of the head and neck over the base of support 
formed by the feet. To us it appears more probable that the curvature is useful to the 
bird when in the water, giving it a greatly increased displacement forward and reducing to 
a minimum the fatigue of carrying the head and heavy beak in the long ocean voyages 
which it undertakes. These may apparently be very prolonged, for the birds have been 
seen no less than 600 miles from land. 

Section II. — Some Peculiarities in the Arrangement of the Fascial Layers 
in Relation to the Pectoral Muscles. 

The most striking feature with regard to the pectoral muscles is their enormous 
size and power. There are four muscles belonging to this group : first, the pectoralis 
major; second, the dermo-humeralis (muscle des parures) ; third, the pectoralis medius ; 
fourth, the pectoralis minor. The origins and insertions of these muscles are given in 
detail and with great accuracy by Professor Morrison Watson. Their action is not 
without interest. The pectoralis major is divided into two distinct parts : an anterior 
which arises from the clavicle and from the outer surface of a strong aponeurosis 
which separates it from the medius and is inserted through a special tendon into the 
whole length of the anterior or radial margin of the bones of the wing, and by an 
aponeurotic extension from the tendon which covers both surfaces of the wing and 
conceals the blood-vessels and nerves, and a posterior which terminates in a V-shaped 



OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 225 

tendon inserted into the humerus. These two portions have different actions ; the 
anterior carries the wing forward to a position at right angles with the trunk, while the 
posterior draws the limb back in the effective part of the swimming stroke, and by its 
insertion into the anterior margin of the humerus it " brings about that rotation of the 
wing round its long axis which, combined with the backward stroke, gives rise to the screw- 
like motion of the organ observable when the bird is progressing through the water."* 

The nerve supply is provided for by a special branch from the large cord of the 
brachial plexus. 

The dermo-humeralis arises from a strong fascia which covers the external oblique 
muscle ; at the posterior margin of the sternum it is reinforced by a number of fibres 
arising from the subcutaneous tissue covering the knee joint. It passes forward parallel 
to the outer border of the pectoralis major to be inserted along with the posterior fibres 
of that muscle into the anterior margin of the humerus. Its action apparently is to 
co-operate with the external part of the pectoralis major in depressing the wing. It is 
supplied by a twig from the nerve to the pectoralis major. 

The 'pectoralis medius passes over a pulley formed by the bones of the shoulder and 
acts as the great elevator of the wing ; by some authors it has been called levator 
humeri. It also is supplied by a twig from the nerve to the pectoralis major. 

The pectoralis minor is inserted into the outer margin of the tricipital fossa ; its 
chief action appears to be to rotate the humerus in a direction opposite to that of the 
rotation produced by the action of pectoralis major. In addition, however, it must help 
in producing the effective, propulsive stroke of the wing. Its nerve of supply is a twig 
from the nerve to pectoralis major. 

The arrangement of the fascia in regard to these muscles is striking and peculiar. 
Tracing the deep fascia of the neck downwards, it passes deep to the clavicle, to which 
it is lightly attached ; and, passing to the thoracic region, it lies superficial to the 
pectoralis medius and minor and becomes continuous with the strong fascia which 
covers the external oblique. Superficial to this fascial layer lie the pectoralis major and 
dermo-humeralis, and from it they both obtain a portion of their origin. The super- 
ficial aspects of pectoralis major and dermo-humeralis are in close contact with skin — so 
close, indeed, that they are marked by regular rows of pits which accommodate the 
rounded ends of the feather quills. 

From this arrangement it appears to us possible that the dermo-humeralis, pectoralis 
major, and clavicle are in reality cutaneous structures. That the dermo-humeralis is in 
series with the dorsal cutaneous muscle and panniculus carnosus is certain ; that it is 
closely associated with the pectoralis major in position, in action, and nerve supply is 
also certain ; but whether the pectoralis major can justly be associated with the same 
group is, on the present evidence, not certain, although it appears to us probable. Un- 
fortunately the evidence from nerve supply is so weakened by the association of 
pectoralis medius and minor with the same nerve as to be valueless. 

* Quoted from Professor Morrison Watson's memoir. 



226 Oil 1). WATERSTON AND DR A. CAMPBELL GEDLES ON EMBRYO PENGUINS 

The suggestion that the pectoralis major is in reality a cutaneous muscle seems to 
us to be not without interest in view of the occasional occurrence in man of the 
abnormal muscle, sternalis. Sir William Turner has suggested that it is to be 
regarded as a vestige of the panniculus carnosus. Professor D. J. Cunningham has 
pointed out that there is considerable evidence to show that it is formed by a devia- 
tion or dislocation of a portion of the pectoralis major ; Mr F. G-. Parsons has shown 
that in rodents the panniculus carnosus possesses two strata, and that there is 
evidence to show that in man the deeper stratum of the panniculus forms the fascial 
sheath over the external oblique and that possibly a portion of the sterno-mastoid is 
derived from the same stratum. He has also shown that there are good reasons for 
regarding the pectoralis major as derived from the panniculus. To these observations 
we now add the facts of the anatomy of the penguin, which appear to us to suggest that 
the apparently antagonistic views of Sir William Turner and Professor Cunningham 
are in reality not opposed, but complementary. 



PART II. 

ON THE EMBRYOLOGY OF THE PENGUIN. 

Introductory. 

As has been shown, the anatomy of the adult penguin has been previously very 
completely described, and little remains to be added to complete our knowledge of it. 

Of the embryology, on the other hand, the existing knowledge is incomplete, owing 
presumably to the great difficulty in bringing back the necessary material in a con- 
dition which allows of a detailed examination being made. A small number of 
embryos was collected by the Challenger Expedition, but the condition in which they 
were received by Professor Morrison Watson rendered them useless for description. 

The material which was put into our hands consisted of a number of specimens each 
one of which had been removed from its egg upon a different day of incubation. It 
would, therefore, appear that we had a specimen for each 24 hours of development. In 
one sense this was so, in another not. The method of collection which was of necessity 
adopted was as follows: — The nest was watched, and the eggs were marked with the 
date of their first appearance and were subsequently collected upon the desired day. 
The difficulties of so doing and the fallacies necessarily attendant upon it are obvious. 
First, it was quite impossible for the observer to know in which of the 24 hours preced- 
ing the marking the egg had been laid ; secondly, it is known to be no uncommon thing 
for the males and females to fight for possession of their egg and for the privilege of 
incubating it. In the course of these struggles the eggs are apt to be dropped and to 
lie for some time directly upon the ice. The result of this must be to retard for a time 
the processes of development, and therefore, although an egg may have been laid for 3, 



OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 227 

6, 9, or 12 days, it does not follow that the development has been proceeding for the 
same number. 

Be this as it may, we found it necessary, especially in dealing with the early 
specimens, to revise the sequence of specimens so as to avoid the absurdity of embryos 
of an obviously earlier stage of development being classified as older than specimens 
of a later stage. The number of days of incubation is, according to the series of 
specimens, thirty. This is surprisingly short, and as there are some obvious gaps 
in the series, we are not prepared to express an opinion as to whether it is correct 
or not. 

The embryos were preserved in 2| per cent, formalin, and we desire to place on 
record our sense of real obligation to the collectors for the care and trouble which they 
took, and for their great skill in packing the specimens, which alone is responsible for 
the excellent state of preservation in which many of these exquisitely delicate specimens 
were when put into our hands. 

As a result of the long time which elapsed between the date of collection and 
examination and of the prolonged immersion of the specimens in formalin, it has not 
been possible to carry out, satisfactorily, any very fine histological examination nor to 
determine with absolute accuracy the exact date of the first appearance of the primary 
ossific centres, the formalin, apparently, having exercised a slight decalcifying action. 
Our investigations have, therefore, of necessity been somewhat restricted, and many 
important problems in connection with the development of the penguin remain unsolved. 

In dealing with the embryology of the penguin the following points seemed to us 
important : — 

1. The enormous length of time through which the penguin has persisted without 
undergoing important modification. In the Quart. Journ. Geol. Soc, vol. xv. p. 670, 
Professor Huxley described a penguin of large size whose remains were obtained from 
the pliocene strata of New Zealand. For the reception of this bird he established the 
genus Palaseudyptes. 

It is exceedingly doubtful whether the establishment of a new genus was necessary, 
and Professor Morrison Watson in his report on the anatomy of the Spheniscidse 
(Challenger Report, vol. vii. p. 46), expresses the opinion that this pliocene bird 
might very well be regarded as belonging to a species of the existing genus Eudyptes. 

From the outset, therefore, we were prepared to meet with some peculiarities of 
development of a more primitive avian type than that obtaining in modern birds. 

2. The enormously high specialisation undergone by the forelimb of the penguin in 
the course of which it has been transformed from an organ of serial to an organ of sub- 
aqueous flight. 

3. The modification undergone by the cervical spine to secure sufficient water dis- 
placement anteriorly to carry the head and heavy beak without fatigue during the long 
ocean voyages which the penguins undertake. 

4. The skeletal arrangements necessitated by the quasi-erect attitude. 



228 DR D. WATERSTON AND DR A. CAMPBELL GEDDES ON EMBRYO PENGUINS 

Methods of Examination. 

The earlier specimens were prepared for examination by being stained in borax 
carmine or alum carmine and mounted entire. Duplicate specimens for examination 
by section were not available. The later specimens were cleared so far as possible, 
but owing to the long immersion in formalin the clearing did not yield very satisfactory 
results. A figure illustrating the development of the limb bones is included among the 
illustrations (fig. 26). The decalcifying action of the formalin has, unfortunately, 
made any definite statements as to date of appearance and order of appearance of the 
ossific centres impossible, and we have had to content ourselves with the negative 
observation that no unusual or abnormal processes or sequences were to be observed, 
although such were looked for as carefully as the material available would permit. 

In order to carry out a comparison regarding the dates of. acquisition of the external 
adult features and the progressive development of the limbs and trunk during the 
later stages of development, we found it necessary to provide a control by carrying out 
observations on the same lines upon the embryo of another genus of birds. 

Data for this purpose were not available in the literature, and we therefore prepared 
and examined, by methods identical with those used for the penguins, a series of duck 
embryos at every 24 hours of incubation. The period of incubation in the duck (28- 
30 days) approximates sufficiently closely to that of the penguin (30-32 days ?) to 
enable comparisons to be drawn with accuracy between embryos of corresponding age 
taken at each 24 hours. 

The data which we obtained appear to be new, and we have, therefore, included an 
account of the external form and the measurements of the embryo-duck with those of 
the penguins. 



Section I. — Early Stages of Development. 

EMBRYOS SHOWING EARLY STAGES, COMPRISING SPECIMENS UP TO THE TWELFTH DAY. 

Specimen I. — In the first of these, which was the youngest specimen examined, 
there was an oval area pellucida measuring 3 mm. in length on the surface of the 
blastoderm. This specimen illustrates the earliest stage after the close of segmentation 
before the formation of the primitive streak, and it closely resembles in size and shape 
the corresponding area in the ovum of the chick and the duck. No area opaca could be 
made out, nor was there any sign of the embryonic shield. The shape of the area 
pellucida is comparable to that seen in the blastoderm of the chick at about 1 8 hours, 
but in the absence of primitive streak and embryonic shield it resembles the chick 
blastoderm at the commencement of segmentation. 

Specimen II. (fig. 1). — The second specimen showed a more advanced stage. The 
germinal area was somewhat pear-shaped. At its wider end was a narrow, semi- 



OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 229 

circular, deeply marked line running round the periphery of the clear area. This line 
appeared to be the commencement of the amniotic fold. Behind this line was a 
narrow area in which there were a few dark spots, which marked apparently an early 
stage of the formation of a vascular area. This area, in turn, merged into a crescentic 
clear area which was limited on the opposite side by a short crescentic dark line 
parallel to the first line. This line occupied the central part of the wider end of the 
clear area, and lay across it, and it appeared to be the rudiment of a commencing head 
fold. From the concavity of this fold a darker streak passed for a short distance 
backwards towards the narrow part of the clear area, and this structure appeared to us 
to be an indefinite stage in the formation of the primitive streak. 

Specimen III (fig. 2). — This specimen resembled that of the chick of 22 hours. 
The head fold was well formed, and behind it lay the medullary folds. These diverged 
from one another at their posterior extremity, and behind them lay the remains of the 
primitive streak stretching to the posterior end of the clear germinal area. The 
primitive streak had the form of two narrow dark lines, enclosing between them a 
narrow clear streak. 

Specimen IV. (fig. 3). — Labelled as 3 days. — The embryo was now 3 mm. in length 
and presented well-formed medullary folds, a notochord, seven pairs of somites, and also 
the remains of the primitive streak which was separated by a narrow clear area from the 
posterior end of the notochord and extended for a distance of 1 mm. behind it. The 
appearances were very similar to those of the chick at 25 hours. The medullary folds 
were as yet ununited, but showed at the cephalic end evidence of distinct thickening, 
while at the posterior ends they diverged from one another in a V-shaped manner. The 
notochord could be traced as far forward as the cephalic enlargements (but this part of 
the embryo was rather damaged), while posteriorly it terminated in a club-shaped 
enlargement. 

Specimen V. (fig. 4). — The succeeding specimen, labelled as being 5 days, 
showed a slightly more advanced stage of the same condition, closely similar to the 
chick at 26 hours. 

The embryo measured 3 "5 mm. in length, and nine pairs of somites were present. 
The posterior end of the notochord was enlarged and shaped like an Indian club, and 
extended beyond the termination of the medullary folds, which diverged outwards on 
either side of it. Behind this extremity was a clear area, and behind it again lay two 
short parallel dark lines, the representatives of the primitive streak. 

The cephalic ends of the medullary folds were considerably enlarged, to form the 
cerebral vesicles. These vesicles were not distinctly marked off from one another, but 
the anterior part was bent forwards, to form the earliest cephalic flexure. In front of 
this again was a narrow clear crescentic area — the pro-amniotic region. 

Specimen VI (fig. 5). — Labelled 6 days. — The next specimen was somewhat 
larger, measuring 6 mm. in length. Twelve pairs of somites were present, and the 
neural folds were larger, and had apparently united with one another at the fourth 



230 DR D. WATERSTON AND DR A. CAMPBELL GEDDES ON EMBRYO PENGUINS 

somite. The folds approached one another closely at the posterior end, and in that 
region they enclosed a clear area, shaped like an arrow head, in which lay the expanded 
posterior end of the notochord. The end of the notochord projected backwards beyond 
the medullary folds in the form of a bulbous extremity. Behind it there was no 
distinct evidence of a primitive streak. The whole of the posterior portion was 
enclosed in a clear oval area. The somites were well formed, and their internal 
structure showed them to consist of a clear central portion and a periphery or 
cortex arranged in columns radiating outwards. 

At the outer margin of the somites was a clear area in which lay a narrow darker 
line, extending along the side from the third anterior somite backwards to beyond the 
hinder somites. 

In position this structure corresponds to the primitive cardinal vein. 

The neural folds showed evidence of segmentation, being constricted opposite the 
centre of the somites, and they were united about the middle. The specimen described 
was one said to be of the sixth day, and the subsequent specimens, which showed stages 
of gradual advance, were in harmony with this statement. 

If these dates be accepted, it is obvious that the early changes in the penguin 
embryo are very much slower than the corresponding changes in the chick, for the 
same degree of development is shown in the chick of 40-44 hours. 

Specimens VII., VIII, IX., and X. (fig. 6). — Labelled 7-11 days. — These 
specimens showed that the next stages of development are very similar to 
corresponding stages in the development of the chick, and do not require detailed 
description, except in so far as they show differences from it. The head and trunk 
show gradual increase in length and in thickness. 

At the hinder end of the notochord a dark area shaped like a spear-head persists 
for some time, but eventually becomes merged in the hinder end of the trunk. 

Figs. 7, 8, and 9, which illustrate the development of the duck at 5, 6, and 7 days, 
are introduced for comparison with fig. 6, which is a photograph of a penguin embryo 
labelled as 8 days. 

Section II. — Kate of Growth of the Trunk and Limbs from the 

Twelfth Day onwards. 

Our observations were directed first to an examination and measurement of the 
developing trunk and limbs. 

The observations begin with the twelfth day of incubation, which is the date of 
appearance of the limb buds upon the surface of the body. 

The details of the measurements are recorded in the appended table (Table 
No. I.). 



OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 231 

Notes upon and Comparison of the Measurements. 
The measurements were analysed along several different lines. 

1. General Growth of the Trunk. 

A. Penguin. — The vertex to coccyx length of the specimen of 12 days (fig. 10), taken 
as it lay in the egg, i.e. in a flexed position, was 15 "5 mm., while that of the duck of 
corresponding age was practically twice as great, namely, 30*6 mm. In the penguin 
the rate of growth from this date onwards was fairly uniform and progressive, and the 
largest embryo, that of 30 days, measured 61*2 mm. in the same position. The 
specimen labelled as being 24 days was larger than those of 25 and 26 days, and 
among the other specimens similar anomalies were present. Taking the maximum 
measurements of specimens which seemed to be typical in their rate of growth, we 
found that the length attained at 12 days was doubled at 19 days and doubled again at 
27 days, the progression being thus almost in geometric ratio. 

The same ratio is found to exist when measurements of the breadth of the specimen 
are taken. The maximum breadth attained at 12 days was rather more than doubled 
at 19 days, and rather less than doubled again at 27 days. 

B. Duck, — The maximum length of the embryos of this species was uniformly 
greater than that of the penguin of same age. Even the largest fully developed 
penguin embryo was smaller than the largest duck embryo. At 12 days (fig. 20) the 
duck embryo was almost exactly twice as long as that of the penguin of corresponding 
age. The rate of growth from this date onwards, as ascertained by the same criteria 
as in the case of the penguin, was, however, much slower than in that creature. 

At 19 clays the length had increased only by rather less than one-half, i.e., was 
4'36 cms. as against 3"06 cms., and not until the 23rd day of incubation was the length 
doubled, and from this time until the end of incubation the increase again was rather 
more than half, from 43*6 mm. to 76'8 mm. The breadth figures showed a rate of 
growth very similar to that found in the penguin. The breadth at 12 days was 10 mm., 
and this figure was practically doubled at the 19th day and almost doubled again at 
the end of incubation. 

2. Rate of Growth of Different Segments of the Trunk. 

Penguin. — Analysing the rate of growth of the different segments which make up 
the vertex to tail measurement, the head segment is found to increase slowly at first, 
but grows rapidly in the last few days, while the neck increases rapidly through all 
stages. 

The growth of the trunk and tail follows the general rate of the whole body, the 
length attained at 12 days being doubled at 19 days, and doubled again at 27-28 days. 

The figures for the head are modified by the fact that measurements were taken 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART II. (NO. 10). 34 



232 DR D. WATERSTON AND DR A. CAMPBELL GEDDES ON EMBRYO PENGUINS 



Table 







Days 12 






14 


15 


16 


17 


Adelia. 






13 


18 


18 




Vertei to coccyx length, Hexed 


P. 


D. 


P. 


D. 


P. 


D. 


P. 


D. 


P. 


D. 


P. 


D. 


P. 


D. 


P. 


D. 




1-55 


$•06 


1-94 


3-36 


1-03 


3 -32 


1-94 


4-07 


2'42 


3 95 


2-58 


4 


335 




2-86 


4-45 




Maximum breadth . . . . 


•64 


V28 


*l-5 


1-41 


*-89 


1-6 


*-91 


1-5 


1-09 


i-s 


1-53 


1-8 


1-46 




1-35 


2-42 




Length of head ..... 


•76 


1-47 


•34 


1-24 


1-01 


1-3 


•94 


1-34 


1-05 


V23 


1-49 


1-6 


1-25 




1-21 


2-1 




,, neck ..... 


•48 


1'4 


•56 


1-65 


•68 


1-5 


•63 


1-24 


•67 


1-32 


•82 


1-35 


1-06 




1-28 


1-74 




„ trunk, including tail 


1-32 


2 '22 


1-44 


2-65 


1-41 


S-09 


1-35 


3-55 


2-01 


3-74 


2'13 


3 '82 


2-71 




2 '42 


4'45 




Cloaca to furcal angle .... 


•78 


1-5 


1-08 


1-82 


114 


1-92 


•98 


2-12 


136 


2-5 


1-44 


2-45 


1-92 




1-65 


3-29 






Tij) of beak to occiput 


•78 


2-05 


1-21 


2-38 


1-29 


2-6 


•98 


2-63 


1-68 


2-86 


1-93 


3-96 


2-28 




2 '06 


3-23 






,, ,, to ear 


•52 


1-59 


•7 


1-82 


•82 


2-04 


•7 


2-1 


1-13 


2-34 


1-43 


2-3 


176 




1 39 


2-7 






,, ,, to post angle of eye . 


•63 


1-44 


•86 


V64 


•84 


1-7 


•74 


1-87 


•97 


V97 


133 


2 


1-53 




1-3 


2-11 




Head^ 


,, ,, to mouth 
,, ,, to nostril 


•25 


■7 


•43 


■92 
•6 


•29 
No 

N. 


1-05 
■7 


•29 
No 

N. 


1-1 
■64 


•61 
No 

N. 


1-2 
•74 


1-78 


1-17 
■8 


•99 

No 

N. 




•78 
No 
N. 


1-23 
■87 






Breadth between eyes 


•64 


1-23 


1-5 


1-34 


•89 


135 


•91 


V35 


1 09 


V35 


1-33 


1-45 


1-46 




1-35 


1-64 






,, ,, ears 


•35 


■84 


■55 


■74 


•46 


■9 


•54 


1-13 


•08 


1-07 


•87 


VI 


•97 




1-04 


1-5 




( Length ..... 
Neck \ 


•48 


1-4 


•56 


1-65 


•68 


1-5 


•63 


1-24 


•67 


1-32 


•82 


1-35 


1-06 




1-28 


1-74 




(Width 


•31 


•S3 


•43 


■45 


•38 


■44 


•31 


■57 


•37 


■73 


•54 


■78 


•56 




•53 


■9 






' Upper arm, anterior border 


•11 


■44 


•19 


■44 


•23 


■46 


•18 


■5 


■37 


■56 


•35 


■55 


•54 




•35 


■64 






,, ,, posterior ,, . " . 




•39 


11 


■61 


•15 


■57 


•12 


■64 


•21 


•7 


•26 


■7 


•33 




•28 


■75 




Fore- 
limb H 


Forearm, anterior border . 
,, posterior ,, 

Elbow, width .... 


•29 
incl. 
Hand 

•13 


•37 
■57 

•31 


•41 

•58 
incl. 
Hand 

•16 


■38 
•44 


•45 
f52 

•17 


■47 
■62 

■53 


f41 
t-48 

•16 


■6 
■82 

■53 


t79 
•84 

•24 


■6 

■78 

■57 


t-91 
1-1 

•26 


•73 
•8 

■55 


tl'21 
1-25 

•28 




fl-04 
tl-19 

•31 


•75 
•87 

■63 






,, thickness 


•08 


■19 


09 


•25 


•07 


•22 


•08 


•26 


•09 


■12 


•09 


■27 


0-9 




•09 


■28 






Length of hand 




•72 




■72 




■86 




•9 




1-21 




V3 








123 






Width „ „ . 




•19 




■21 




•18 




•19 




•2 




■32 








■35 






. Length of pollex 




•18 




•18 




■34 




•24 




■35 


... 


■37 








■23 




Leg, anterior border .... 


J'54 


1-04 


t-72 


■98 


t-73 


1-05 


J-61 


1-03 


•57 


1-23 


•55 


V35 


•77 




•86 


1-55 




,, posterior ,, . . . . 




■53 




■42 




■52 




■74 


•21 


•82 


•18 


•86 


•28 




•22 


1-12 






Plantar, heel to end of tliird 

phalanx 
Plantar, length of phalanges 




1-16 

•55 




1-45 
■65 




V52 

■73 




V65 

■75 


•54 
•28 


1-97 . 
■96 


•69 

•28 


2-08 
1-15 


•91 
•45 




•78 
•37 


2-57 
V23 




Foot • 


Plantar, web between outer and 

middle 
Plantar, web between inner and 

Illi'Ml': 

Dorsal, ankle to tip of nail of 

middle t*>e 
Dorsal, length of nail 




1-04 




41 
1-3 




■42 

■45 

1-35 




■47 
•4 
1-6 

■1 


•18 
•12 
•64 


'5 

•64 

1'82 

■14 


•17 
■14 
•54 


■7 

■65 

1-94 

■13 


•19 
•14 
•82 
•1 




•17 
•12 
•78 
•09 


■75 

■63 

2-25 

•28 





Over eyes. 



t Elbow to tip. 



Knee to tip. 



OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 



233 



19 



20 



P. 


D. 


2-99 


4' 36 


1-45 


2- 35 


1-2 


1-3 


1-09 


2-05 


27 


4 


184 


2-6 


2-34 


3-1 


17 


2S4 


1-44 


2 


•92 

No 

N. 


1-25 
■83 


1-45 


1-45 


•98 


1-22 


1-09 


2-05 


•7 


■7 


■46 


•69 


•32 


■82 


tl-23 


■76 


T34 


1-07 


•32 


■53 


■1 


■27 




1-5 




•26 




•38 


•89 


1-18 


•3 


1 


•89 


2-45 


•43 


1-14 


•29 


'85 


•21 


■8 


•82 


2-2 


•11 


•26 



P. 



3-37 
T52 
1-18 
•99 
32 
2-24 
2-43 
1-89 
1-59 

1-1 

No 

N. 

1-52 

1-15 

•99 

•77 

•51 

•45 

tl'34 

1-44 

•35 

•11 



D. 



•91 
•4 
1-12 
•54 
•34 
•22 
•93 
•14 



6-25 
2-73 
1-2 
1-94 
5-25 
4-16 
3-45 
2-78 
2S 
V43 
•91 
1-54 

1-4 
1-94 

1 

1 

•96 

■97 

1-23 

•6 

•23 

V5 
■42 
■43 

2-3 

1-54 

3-56 

1-9 

1-17 

1-16 

29 



21 



P. 



D. 



5-93 

2-95 

1-43 

195 

5 01 

3-46 

5S3 

2-73 

243 

1-41 

■97 

1-57 

145 

1-95 

VI 

■53 

■64 

■93 

1-07 

■52 

■25 

1-55 

•4 

•5 
2-15 
V4 
3-24 
1-47 
1-18 
1-18 
3-07 



22 



P. 



3-49 



D. 



5-05 

2-8 

1-55 

2-15 

5-4 

3-13 

3-64 

2-75 

2-25 

1-43 

■9 

1-53 

V43 

2-15 

1 

■5 

■83 

•64 

■96 

■51 

■23 

1-3 

•37 

•21 

2-3 

1-68 

3-42 

V55 

1-25 

1-25 

3-23 

•35 



23 



P. 



4-09 

1-89 

1-48 

1-66 

4-06 

3*21 

2-81 

2-26 

1-85 

1 29 

•57 

1-7 

1-27 

1-66 

•8 

•77 

•47 

tl'65 

tl'72 

•48 
•16 



1-21 
•48 

1-25 
•61 
•34 
•23 

1-34 
•25 



D. 



6-39 

3-22 

1-97 

234 

6-02 

4-56 

3 9 

2-9 

2S 

1-58 

■94 
1-6 
1-52 
2-34 
1-08 

•78 
1 

1-1 4 
1-07 

■56 

•26 

1-46 

■43 

■56 

2-53 

1-55 

4-2 

V93 

1-46 

1-46 

3-62 



24 



25 



P. 



4-64 

2-24 

1-47 

1-63 

4-16 

3-41 

3-2 

2-57 

1-93 

1-5 

•58 

1-88 

1-54 

1-63 

1-09 

1-11 

•57 

tl-87 

t2-07 

•56 
•14 



1-6 

•46 
1-73 

•8 

•55 

•31 
1-53 

•29 



D. 



7-04 
3S5 



6 

4-41 

4-02 

2-95 

2-47 

1-6 

1 

1-77 

1-66 

2-62 

1-05 

■56 

■8 

■9 
VI 



1-75 
•52 
■73 
2-8 
2-22 
3 93 
V73 
1-54 

1-54 

3-64 

•44 



D. 



6-06 

3-8 

2S5 

2-93 

5-83 

4-61 

4-01 

3 

2-56 

1-66 

1-02 

1-58 

1-54 

2-93 

1 

■64 
1-27 
1-08 
V23 

■55 

■29 

1-75 

•34 

■64 

2-96 

V88 

4-23 

2-03 

V52 

1-52 

3-62 



26 



P. 



D. 



7 68 
4-46 
1-95 
2-88 
6S2 
4-4 
4-4 
3-27 
2-87 
V82 
1-18 
V64 
1-73 
2-88 
V2 
■74 
V43 
1-18 
V32 

74 

•39 

V83 

■52 

■64 

3-18 

2-14 

4'61 
2-03 
1-7 
1-72 
4-1 
■47 



27 



28? 



30? 



I). 



6-07 

2-35 

1-91 

1-92 

4-91 

4-15 

3-43 

272 

2-07 

1-66 

•59 

1-87 

1-69 

1-93 

1-23 

•79 

•58 

t2-17 

t2-13 

•64 
•22 



1-61 
•66 

1-85 
•93 
•65 
•41 

1-79 
•39 



P. 

5-71 
3'05 
2-55 
1-85 
5-51 

3-98 
3-07 
2-39 
1-91 

•7 

2-02 

1-83 

1'85 

1-25 

•77 

•81 

•68 

1-01 

•76 

■23 

1-38 



2-18 

•73 

2-32 

1-01 
•81 
•53 

2-02 
•46 



6-12 
3 03 
2-89 
1-92 
5-35 

421 

3-28 

2-56 

2-01 

■88 

1-96 

1-89 

1-92 

1-31 

•93 

•96 

•68 

•91 

•83 
•36 
1-5 



2-12 
•66 

2-43 

1-07 
•83 
•52 

2-3 
•41 



Over eyes. 



t Elbow to tip. 



Knee to tip. 



234 DR D. WATERSTON AND DR A. CAMPBELL GEDDES ON EMBRYO PENGUINS 

from the tip of the beak, and consequently the increase in total length is found to be 
greatest in the last few days, when the beak elongates rapidly, while the figures 
obtained for the growth of the head in width show a more gradual and regular increase, 
comparable to the rate of growth of the trunk. 

3. Growth of the Limbs. 

A. Penguin. — Fore Limbs. — The first appearance of the fore limb had the form 
of a bud-like projection from the lateral aspect of the trunk (fig. 10). This process at 
first was somewhat spatulate, and then rather bulbous at its free end (figs. 11 and 13). 
This shape was in turn soon lost, and the limb rudiment assumed the characteristic 
sickle-shaped form found in the adult bird (figs. 15 and 17, 19 and 23). The two 
segments of the arm were soon clearly to be recognised and were separately measured. 

The measurements taken were those of the anterior border of the upper segment, 
together with the anterior border of the forearm from the elbow to the tip. At 1 2 days 
the total length was 4 mm., made up of l'l mm. of upper segment and 2*9 mm. of the 
lower. 

At 19 days the length was 13*9 mm. and at 27 days 29'7 mm. The rate of growth, 
therefore, had been such that the length of the limb was more than trebled in the first 
period, and was more than doubled in the second. Both segments shared the growth, 
and did so in fairly equal proportions throughout. 

Lower Limb. — The portion of the lower limb which was free from the coverings of 
the trunk was measured in one or in two pieces, as the size permitted. At 12 days 
it measured 5 "4 mm., at 19 days 18"1 mm., and at 27 days 36*2 mm. In the first 
period the length was more than trebled, and in the second it was again doubled, 
showing a rate of growth almost exactly similar to that found in the upper limb. 

B. Duck. — In the duck corresponding measurements showed that at 12 days (fig. 20) 
the fore limb measured 15*3 mm. and was nearly four times as long as in the penguin. 
At 19 days the length was almost doubled (28*5 mm.), and at the end of incubation it 
had only added one-third to its length (37 '5 mm.). The lower limb, measured in a 
similar way at 12 days, was 22 mm. in length, again four times as long as in the 
penguin, and in the specimen of 20 days it was more than doubled, and at the end of 
incubation it was almost doubled again. 

If the proportion which the length of the limbs bears to the total length of the 
body be examined, it is found that in the penguin the fore limb at 12 days is almost 
one-fourth of the body length, that it increases rapidly for a few days until it measures 
marly one-half of the length of the body, and maintains approximately this ratio till 
near the end of the incubation period, when it progresses rapidly and assumes a ratio 
of rather more than two-thirds of the body length. 

In the duck the ratio at 22 days was one-half, and this proportion became rather 
larger, but eventually again became nearly one-half. 



OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 



235 



Table II. — Vertex to Coccyx Length, the Length of Limbs and the Ratio between them, 

the Former being taken as 1. 





Vertex to Coccyx 


Length of Upper 


Index- 


Length of Lower 


Index. 






Length. 


Limb. 






Limb. 






Days. 
















Penguin. 


Duck. 


Penguin. 


Duck. Per 


iguin. 


Duck. 


Penguin. 


Duck. 


Penguin. Duck. 


12 


1-55 


8-06 


•40 


1S3 


25 


•50 


•54 


2-20 


•34 


71 


13 


1-94 


3-36 


•60 


1-54 


30 


•45 


•72 


2-43 


•37 


72 


14 


1-93 


3-32 


•68 


1-79 


35 


•53 


•73 


2-57 


•37 


77 


15 


1-94 


4-07 


•59 


2 


30 


•49 


•61 


2-68 


•31 


65 


16 


2-42 


3-95 


116 


2-37 


47 


•60 


1-11 


3-20 


•45 


80 


17 


2-58 


4 


1-26 


2-58 


48 


•64 


1-26 


3-43 


•48 


85 


18 


2-85 


4'45 


1-75 


2 62 


61 


•58 


1-71 


4-12 


■60 


92 


19 


299 


4-36 


1-39 


2-95 


46 


•67 


1-81 


363 


•60 


83 


20 


337 


6-25 


1-85 




54 




1-84 


5S6 


•54 


85 


21 


3-62 


5-93 


2-16 


3-01 


59 


•50 


2-42 


5-39 


•66 


90 


22 


349 


5-05 


2-10 


2-24 


60 


•44 


- 2-38 


5-72 


•68 1 


!3 


23 


4-09 


6-39 


2-42 


3S8 


59 


•52 


2-62 


6-73 


•63 1 


05 


24 


4-64 


7-04 


2-98 


3 21 


64 


•45 


352 


7-73 


•76 1 


09 


25 


4-13 


6-06 




3'47 




•57 




7-19 


1 


18 


26 


3-99 


7-68 


2-45 


3-75 


61 


•48 


3-21 


779 


•80 1 


01 


27 


6-07 




2-97 




48 




3-62 




•59 




28 


571 




3-53 




61 




4-71 


... 


•82 




30 


612 




4-24 




69 


... 


4-79 


... 


•78 





Synopsis of Results from Comparison of the Measurements of the Penguin 

and Duck. 

1. Length of Trunk. 

At early stages the length of the trunk of the penguin is much less than that of 
the duck of corresponding age — the proportion being that the penguin at the 12th 
day is almost exactly one-half of the length of the duck at the same age. 

As development proceeds the penguin grows more rapidly, until at the end of 
incubation the penguin is only fractionally shorter than the duck. 



2. Length of Fore Limb. 

At corresponding dates after the limbs first appear the fore limb of the penguin is 
only about one-fourth of the length of that of the duck. 

Its growth subsequently is much more rapid, and at the three weeks it is almost 
the same size, and at the end of incubation it exceeds in length the duck's fore limb, 
in some specimens by as much as one-quarter. 

3. The Ratio of Length of Upper Limb to Trunk. 
At the 12th day of incubation the fore limb of the penguin measures approxi- 
mately in length one-fourth (index, '25) of the length of the trunk, while in the duck 



236 DR D. WATERSTON AND DR A. CAMPRELL GEDDES ON EMRRYO PENGUINS 

the limb is equal to one-half of the length of the trunk (index, # 50). The ratio of 
growth of the fore limb in the penguin is continuous and progressive in relation to the 
ratio of growth of the trunk. 

Towards the end of the third week it measures one-half of the length of the trunk 
(index, 21st day, "50), and at the end of incubation it measures two-thirds (index, *69) of 
the length of the trunk. 

In the duck, on the other hand, the length at first increases relatively to the trunk, 
attaining in one specimen, at the end of the third week, a maximum ratio of over two- 
thirds (index, 19th day, "67) ; but thereafter it slowly loses ground, and at the close of 
incubation the fore limb is less than one-half the length of the trunk. 



"&* 



4. Hind Limb. 

A. Absolute Length. — In the penguin at the 12th day of incubation the hind 
limb is also one-fourth of the length of the hind limb of the duck, the ratio being very 
similar to that between the upper limbs of the two animals. In the penguin's limb, 
however, growth is more rapid, with the result that at the end of the third week the 
hind limb has reached a length equal to nearly one-half of the length of the hind 
limb of the duck of corresponding age, and at the end of incubation it is rather more 
than half as long as the corresponding part in the duck. 

5. Ratio of Length of Lower Limb to that of the Trunk. 

At the 12th day in the penguin the length of the lower limb is approximately 
one-third of the length of the trunk (index, *34), but the rate of growth is more rapid, so 
that at the end of incubation it is nearly as long as the trunk (index, 28th day, - 82). 

In the duck the limb at 12 days is rather more than two-thirds of the length of 
the trunk (index, '71), at the 18th day it is nearly as long as the trunk (index, '92), 
and thereafter it continues to increase at a rate relatively rather greater than the trunk, 
so that at the end of the third week it becomes actually longer (index, 22nd day, 1'13), 
and retains this superiority to the end of incubation. 

In this respect, i.e. in the rate of growth from the 18th day, it is the exact reverse 
of the fore limb — it continues to increase in length relatively to the trunk, whereas the 
fore limb actually shows a relative decrease. 

Analysis op Results. 

Among the four limbs examined there are, therefore, two types of development. 

First, there is a progressive type, and to this the development of the fore and hind 
limbs of the penguin and of the hind limb of the duck belongs. 

Second, there is a partially regressive type, and to this the development of the fore 
limb of the duck belongs. 



OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 237 

Now, if it be true that ontogeny is an abbreviated recapitulation of phylogeny we 
are forced to conclude that the fore limb of the penguin is a progressive structure, the 
fore limb of the duck, partially at least, a regressive. The reason for this, doubtless, 
is that the mesoblastic portion of the duck's wing has, with the advancing specialisation 
and adaptation of the epiblastic structures, become less valuable. To all flying birds 
feathers, not bone and muscle, are of prime importance in securing wing spread for 
aerial flight : for the penguin, on the other hand, a rigid bony paddle has been evolved, 
adapted to resist the pressure of the water to which it is subjected when the bird's 
great pectoral muscles are in strenuous action. 

And we thus are led to the interesting conclusions : — 

1. That the mesoblastic structures in the penguin's "wing" are progressive, the 
epiblastic regressive. 

2. That the mesoblastic structures in the duck's wing are regressive, the epiblastic 
progressive : for in the development of the duck we find that the developmental energy 
is suddenly on the 17th day switched off, as it were, from the mesoblastic structures on 
to the epiblastic ; for from that day onward the down rapidly develops and the meso- 
blastic framework loses ground. 

. With regard to these developmental facts the question arises : — 
Is the duck's or the penguin's wing the more direct descendant of the common 
ancestor ; or have they both diverged from the common stock approximately equally, 
but in opposite directions 1 

Embryology alone cannot answer this question, but the evidence is clear in this, 
that the fore limb of the penguin in its development goes through a progressive and 
continuous series of stages along one unbroken line. The mesoblastic portion of the 
fore limb elongates, but its characters do not alter. It elongates, however, with a 
relatively greater rapidity towards the end of development, whereas the duck's fore limb, 
after being relatively longer than the penguin's ever is, regresses rapidly. So that the 
answer to our question, so far as the embryological evidence is concerned, must be that 
the wings of both these birds are different from the ancestral wing, and that the 
differentiation has been in opposite directions and that the common ancestor was a 
flying bird of a somewhat primitive type depending in large measure for the spread of 
its wing upon bone and muscle. 



Section III. — External Configuration of the Embryos from the 

Twelfth Day onwards. 

1. Gentoo Penguin Embryo, 12 Days. (Fig. 10.) 

The embryo was not fully formed, the brain was covered by the thinnest of mem- 
branes, the beak was short and soft, the eyes were very prominent, with dense white 
opaque pupils. The fore limb well developed ; the elbow flexure completely marked, 



238 DR D. WATERSTON AND DR A. CAMPBELL GEDDES ON EMBRYO PENGUINS 

It ended in a somewhat spatula-shaped extremity. The hind limb formed a spatula- 
shaped protrusion. 

Duck Embryo, 12 Days. (Figs. 16 and 20.) 

The embryo was fully developed, the integument was strongly marked, and there 
were down papillae on two short lines at either side of the tail. The eye prominences 
were very marked ; the head and beak were soft and pliable except at the point of the 
beak, where there was a small white nodule of keratin. The fore limb showed distinct 
division into upper arm, forearm and hand. The pollex was prominent. 

2. Gentoo Penguin Embryo, 13 Days. (Fig. 11.) 

The embryo was not fully formed. The head was soft and the brain showed 
through it. The beak was short, the eyes were prominent, the fore limbs were short, 
but showed plainly the elbow joint, and had a flattened, spade-like tip : to the 
naked eye there was no sign of digitation. The hind limbs were short, the ankle was 
unformed, the end was spatular and showed signs of three digits, the tail was relatively 
long and was much curved ventrally. 

Duck Embryo, 13 Days. (Fig. 18.) 

The embryo was fully formed. Down was appearing in lines on the back and sides 
of the embryo, and laterally on the neck ; otherwise as for 1 2 days. 

3. Gentoo Penguin Embryo, 14 Days. 

The embryo was not fully formed, the head was soft, the brain showing through it. 
The beak was short, the eyes prominent ; the fore limbs short, but well formed ; the 
hind limbs were short, spatular protrusions. The heart showed distinctly, shining 
through the thin anterior wall of the thorax. The tail was relatively long and much 
curved ventrally. 

Duck Embryo, 14 Days. (Fig. 21.) 

Down papillae were seen all over the embryo, especially along the dorsal tracts and 
on the tail. The eyes were widely open. 

4. Gentoo Penguin Embryo, 15 Days. (Fig. 13.) 
The embryo was small, the beak was beginning to grow forward. The fore limbs 
were small and soft, but fully formed. The hind limb showed no division into the 
toes, but ended in flat, spatular extremities. There was no sign of the fourth toe 
independent of the flattened extremity. The tail was relatively long and curved 
ventrally. 

Duck Embryo, 15 Days. (Fig. 22.) 

Down was appearing upon the head along a well-defined tract which looped round 
the eye from behind. 



OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 239 

5. Gentoo Penguin Embryo, 16 Days. (Fig. 14.) 
The embryo was small, the nails were not developed, so far as could be seen on 
naked-eye examination. There was a well-marked ridge down the centre of the 
abdomen and thorax. The toes did not appear beyond the web, but looked more as if 
they were embedded in it. The fourth toe appeared as if the web were continued 
round it. 

Duck Embryo, 1G Days. 
As for 15 days. Down more marked. 

6. Gentoo Penguin Embryo, 17 Days. (Fig. 15.) 

The embryo was in good condition and the fully formed integument was complete. 
The skin was marked by papillae regularly arranged. On either side of the tail there 
were several down fibres varying from 1*7 mm. in length to a length just visible to the 
naked eye. 

Duck Embryo, 17 Days. 

The down was well developed. 

7. Gentoo Penguin Embryo, 18 Days. (Fig. 19.) 
This specimen was somewhat macerated. 

Duck Embryo, 18 Days. 
Embryo completely down-covered. 

8. Adelia Penguin Embryo, 19 Days. 

The embryo was fully formed, the integument was complete. There was a well- 
marked median groove over the abdomen and thorax, and from it on either side 
extended twelve transverse grooves. The body was completely naked except for eight 
fibres on the tail (four on either side of the median line). 

9. Gentoo Penguin Embryo, 19 Days. 

The embryo was fully formed, and the integumentary covering was complete. 
There were no transverse grooves on the abdomen. 

10. Gentoo Penguin Embryo, 20 Days. 

The embryo was fully flexed and completely devoid of down. The integument over 
the abdomen was not quite complete. A bare area in length "28 cm. and in breadth 
at base "07 cm. extended forward in front of the umbilicus. There were no transverse 
grooves*on the abdomen. 

TKANS. KOY. SOC. EDIN, VOL. XLVII. PAET II. (NO. 10). 35 



240 DR D. WATERSTON AND DR A. CAMPBELL GEDDES ON EMBRYO PENGUINS 

11. Gentoo Penguin Embryo, 21 Days. 

The embryo was completely flexed ; the limbs were completely formed. The 
integument over the abdomen was incomplete; a triangular area 1"23 cms. in length, 
•07 cm. in breadth, extended forward in front of the umbilicus and was continued over 
the front of the thorax as a well-marked groove from which on either side extended 
twelve transverse grooves. 

12. Gentoo Penguin Embryo, 22 Days. (Fig. 17.) 

The limbs of the embryo were complete ; the integument over the abdominal 
regions was incomplete. An area 1*04 cms. in length and "09 cm. in breadth at the 
umbilicus, gradually tapering to a point, was undeveloped. The head and body were 
entirely naked. The eye prominences were very marked ; the skull and beak were soft 
and pliable. There were five transverse grooves on the abdomen. Flexion was 
incomplete. 

13. Gentoo Penguin Embryo, 23 Days. 
The embryo was not quite fully developed, the integument being incomplete. On 
the anterior abdominal region — anterior to the umbilicus — a very acute isosceles 
triangle persisted uncovered, the sides of which measured 1"19 cms., the base "12 cm. 
The eye prominences were very marked, the skull and beak were soft and pliable. A 
few short pieces of dark down had developed on the head ; the rest of the body was 
absolutely naked. There were no transverse grooves on the abdomen ; flexion was 
incomplete. 

14. Gentoo Penguin Embryo, 24 Days. (Fig. 23.) 

The embryo was fully developed. The prominence of the eye-balls was well marked 
and the eye-slit was open, 18 cms. long ; its width at its broadest part when stretched 
was '08 cm. The head was covered with dark down, and on the back and body there 
was a sparse covering of light-coloured down considerably more than on the 2 5 -days' 
embryo. The middle line of the front of the abdomen showed a deep groove from which 
extended across the abdomen, on either side, six well-marked transverse grooves. The 
feet and tail had now assumed the adult relations, and formed the peculiar fiat base 
which is characteristic of the adult bird and permits of its adopting the quasi-erect 
attitude habitual to it. 

15. Gentoo Penguin Embryo, 25 Days. (Fig. 24.) 

The embryo was fully formed. The prominence of the eye-balls was well marked. 
The eye-slit was small, the beak was soft, and the head was sparsely covered with short 
black down on the back. The body was fully flexed ; the spine formed a semicircle from 
tail to head. There was a well-marked median groove on the front of the abdomen 
and thorax, from which extended transversely on either side eight well-marked grooves. 



OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 241 

16. Gentoo Penguin Embryo, 26 Days. 
The embryo was fully formed ; the eyes looked large and prominent ; the beak was soft. 
The head was covered with short black down ; the rest of the body was practically naked. 
Along the middle line of the abdomen, from the umbilicus to the root of the neck, was a 
well-marked groove. 

17. Gentoo Penguin Embryo, 27 Days. (Fig. 25.) 
The limbs of the embryo appeared fully formed. The feet were flexed over the front 
of the abdomen, plantar surface outwards. The upper limbs lay close to the side and 
over the abdomen under cover of the feet. The neck was flexed so that the beak lay along 
the right side of the abdomen immediately internal to the right superior extremity. 
Down was plentifully present on the head and back, shading from almost black over the 
head to a light slate grey over the sacral region. The down on the abdomen and anterior 
aspect of the thorax was sparse and light in colour. The middle line of the body in 
front was marked by a very evident ridge in the lower part, flanked on either side by a 
o-roove. These grooves coalesced in the upper part of the abdomen and were continued 
over the thorax as a median depression. This could be traced as high as the root of the 
neck, where it was lost. 

18. Gentoo Penguin Embryo, 28 Days. 

The embryo appeared fully formed and was in a fully flexed position, so that the 
under surface of the mandible was pressed firmly against the thorax and abdomen ; the 
beak lay in a groove formed by the right fore limb and left hind limb. The head was 
covered with a dark grey, almost black, down ; the back with slate coloured, the abdomen 
with white. The back of the head had been flattened by pressure. 

Duck Embryos, 19-28 Days. 

The advance was purely one of body bulk. The amount and directions of growth 
are shown by the measurements of the specimens. (See Table I.) 

19. Gentoo Penguin Embryo, 30 Days. (The specimen in fig. 25 closely 
resembled this older specimen, but was rather smaller.) 

The embryo appeared fully formed and was in a fully flexed position, so that the under 
surface of the left side of the head was pressed against the thorax, the beak lying to the 
right side of the abdomen between the right fore limb and the foot of the left hind limb 
anteriorly and the right hind limb posteriorly. The head and back were covered with 
slate-coloured down, the belly with white down. In the flexed position the greatest 
length was 571 cms., the greatest breadth 3 "05 cms. In the extended position the 
length from the crown of the head to the tip of the tail was 8*41 cms. When measured 
separately the head was 2'55 cms. in length, the neck 1 85 cms. ; the body and tail 



242 DR D. WATERSTON AND DR A. CAMPBELL GEDDES ON EMBRYO PENGUINS 

together 5*5 1 cms. — giving a total length of 9'91 cms. The 1*5 cms. of difference (9'91 
cms. — 8 '41 cms.) is accounted for by the marked spinal curves which immediately asserted 
themse^es on extension, a lumbo-thoracic curve ventrally concave, a cervical curve 
dorsally concave. 

Summary and Conclusion. 

Part I. — Anatomy of the Adult Penguin. 

1. The cervical portion of the vertebral column is possessed of a curve of extra- 
ordinary convexity which causes the vertebral bodies to form a visible bulging on the 
front of the neck. This convexity we regard as useful to the birds by increasing water 
displacement anteriorly and permitting of the head and beak being carried without 
undue fatigue on the long ocean voyages which the birds undertake. The furcal angle 
being occupied by the vertebral bodies, the trachea and oesophagus enter the thorax on 
a plane posterior to the cervical spines. This curvature of the column and arrangement 
of the viscera is permanent and characteristic of the adult bird. 

2. The arrangement of the fascial layers in relation to the pectoral muscles suggests 
that the pectoralis major and clavicle are in reality cutaneous structures. It was found 
impossible from the embryological material at our disposal to prove or to disprove this 
suggestion. 

Part II. — Embryology of the Penguin. 

1. The number of days of incubation is, according to the series of specimens, thirty. 
This is surprisingly short, and as there are some obvious gaps in the series, we are not 
prepared to express an opinion as to whether it is correct or not. 

2. The early stages of development are exactly comparable to the corresponding 
stages of the chick or duck, though, if the dates of our specimens be correct, they take 
place more slowly. 

3. The later stages of development are exactly comparable to the corresponding 
stages of the duck except with regard to the down, the fore limb, and the foot. 

4. The foot of the penguin is, throughout its development, exceedingly clumsy and 
primitive, as is the foot of the adult ; the foot of the duck, from its earliest appearance, 
assumes a lighter and neater appearance, and is adapted as a paddle, which the foot of 
the penguin is not. 

5. The fore limb of the penguin is, in its mesoblastic structures, definitely progressive, 
whereas the mesoblastic portion of the duck's wing is relatively regressive. Both, 
however, have varied from the ancestral type, but their variation is in opposite direc- 
tions : the wing of the penguin has required rigidity to subserve its function as a paddle, 
and it has acquired it, throughout its length. The wing of the duck, on the other hand, 
has required lightness, and it has acquired it — the reduced spread of bone being 
compensated for by an increased development of feathers. 



OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 243 

6. The evidence from embryology is in favour of the early divergence of the 
Aptenodytidse from the main stem of the flying birds, and of the common ancestor of 
both having been a flying bird, though of a clumsy type. 

7. We were unable to verify the facts recorded by other observers in the ossification 
of the limbs, as we were handicapped by the decalcifying action of formalin even in 
weak solutions, so that our observations in this direction could not be considered 
altogether satisfactory. 



ILLUSTRATIONS. 
Plate I. 



Fig. 1. Gentoo penguin, embryonic area, early stage. 

Fig. 2. Gentoo penguin, within first 24 hours, showing head fold, medullary folds, and primitive streak. 
This stage resembles that of the chick at 22 hours. 

Fig. 3. Gentoo penguin, said to be 3 days. This stage of development resembles the chick at 24 hours. 

Fig. 4. Gentoo penguin, said to be at 5 days. The stage of development resembles the chick at 26 hours. 
The photograph shows specially the cephalic flexure, the hinder ends of the medullary folds enclosing the 
bulbous extremity of the notochord, and the remains of the primitive streak. 

Fig. 5. Gentoo penguin, said to be 6 days. Comparable to the chick at 40 hours. The photograph 
shows the bulbous expanded posterior end of the notochord, and evidence of a segmentation of the neural 
folds. 

Fig. 6. Gentoo penguin, about 7 days. 

Plate II. 

Fig. 7. Duck embryo of 5 days. Compare fig. 4 of penguin of same age. 

Fig. 8. Duck embryo of 6 days. Compare fig. 5 of penguin of same age. 

Fig. 9. Duck embryo of 7 days. Compare fig. 6 of penguin of same age. 

Fig. 10. Gentoo penguin embryo of 12 days. Both upper and lower limbs have appeared on the surface 
of the trunk. 

Fig. 11. Gentoo penguin embryo of 13 days. 

Fig. 12. Duck embryo of 8 days. Compare character of limbs with those of penguin of 13 davs 
(fig. 11). 

Figs. 13, 14, and 15. Successive stages of Gentoo penguin embryos at 15, 16 and 17 daysj to illustrate 
stages in the development of the sickle-shaped fore limb of the adult. 

Fig. 16. Duck embryo of 12 days, in situ, to show how the embryonic attitude is maintained, and is not 
dependent upon external pressure. 

Fig. 17. Gentoo penguin embryo of 22 days, to show advancing change in the position of the feet. 
Compare figs. 15 and 23, which show other stages in the development of the peculiar flat base of the adult. 



Plate III. 
Fig. 18. Duck embryo of 13 days. 

Figs. 19 and 20. Gentoo penguin embryo of 18 and duck embryo of 12 days, photographed to the same 
scale, to show altering characters of the limbs, when the trunks are of the same size. 
Fig. 21. Duck embryo of 14 days. 

Fig. 22. Duck embryo of 15 days. Compare with fig. 15 of gentoo penguin embryo of the same a«e 
TRANS. ROY. SOC. EDIN, VOL. XLVII. PART II. (NO. 10). 36 ° 



244 EMBRYO PENGUINS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 

Fig. 23. Gentoo penguin embryo of 24 days. Notice the character of the fore limb, and the attitude of 
the hind limb and tail. 

Fig. 24. Gentoo penguin embryo of 25 days. 

Fig. 25. Gentoo penguin embryo of 27 days, to show the stage of development reached before leaving 
the shell. 

Fig. 26. Photograph of fore limbs of gentoo penguin embryo of 22 days, to show ossification. 

The clear portions of the shafts are composed of bone, and the dark extremities are of cartilage. Both 
extremities of the radius and ulna are cartilaginous. Beyond them is a darkish area, in which are the 
cartilaginous rudiments of the carpus. 

Beyond that region are the metacarpal bones and phalanges. 



frans. lioy. Soc. Edin r - Plate I. Vol. XLVII. 

Drs Waterston and Geddes. — Embryo Penguins : Scottish National Antarctic Expedition. 





Fig. 4. 




Fig. 5. 



Fig. 6. 



M'F:irl:me & Erskine, Edin. 



Trans. Roy. Soc. Edin r Plate II. Vol. XLVII. 

Drs Waterston and Geddes. — Embryo Penguins : Scottish National Antarctic Expedition. 





Fig. 7. 



Fig. 8. 





Fig. 11. 



Fig. 12. 






Fig. 9. 



Fig. 10. 





Fig. 13. 



Fig. 14. 



Fig. 15. 





Fig. 16. 



Fig. 17. 



M'Farlane & Erekine, Edin. 



Crans. Roy. Soc. Edin r Plate III. Vol. XLVII. 

Drs Waterston and Geddes.— Embryo Penguins : Scottish National Antarctic Expedition. 




Fig. 21. 




Fig. 24. 




Fig. 19. 




Fig. 22. 




Fig. 25. 




Fig. 20. 




Fig. 23. 




Fig. 26. 



MTarlane & Erskine, Edin. 



( 245 ) 



XI. — The Pharmacological Action of Harmaline. By James A. Gunn, M.A., M.D., 
D.Sc. (From the Pharmacology Laboratory of the University of Edinburgh.) 
Communicated by Sir Thomas R. Fraser, M.D., F.R.S. 

(MS. received November 4, 1909. Read November 22, 1909. Issued separately December 31, 1909.) 

CONTENTS. 



PAGE 

Introductory . . . . . . . 245 

A. Lethality of Harmaline 247 

B. Symptoms produced by Harmaline — 

(a) In Frogs 249 

(U) In Mammals — 

1. Guinea-pigs 250 

2. Cats 251 

C. Action on the Cerebro-Spinal Nervous 

System — 

(a) Brain and Spinal Cord .... 252 

(6) Nerve 254 



PAGE 

D. Action on Skeletal Muscle .... 255 

E. Action on the Circulation — 

(a) Heart 256 

(6) Blood-vessels 260 

(c) Heart and Blood-vessels (Blood Pressure) 261 

F. Action on Respiration ..... 269 

G. Action on Temperature 269 

H. Action on the Uterus . . • . . . 270 

General Summary . . . . . . 271 



Introductory. 

Harmaline is one of two alkaloids found in the seeds of Peganum Harmala, a 
strong-smelling herbaceous plant belonging to the order of Rutaceae. This plant 
grows to a height of from 1 to 3 feet, is much branched, and profusely covered with 
leaves. It is found wild in S. Europe, Asia Minor, Egypt, Arabia, N.W. India, and 
Siberia. It is the Hnyavov aypiov (wild rue) of Dioscorides, Htyavov being the name 
still applied in Greece to several species of Ruta. 

The seeds were used medicinally by the ancient Greeks, as they are to this day in 
India, where they are known chiefly by the old Arabic name of Harmal. In Europe 
they were formerly much employed as Semen Rutce sylvestris, and as such are enu- 
merated among the simples of several of the early London Pharmacopoeias.* 

The seeds are of a dark brown colour, and contain (1) a red colouring matter, 
which was at one time imported into England from the Crimea as a dye ; (2) oil ; 
(3) a soft resin of a deep carmine lake colour, having a heavy odour like that of 
Cannabis Indica ; (4) two alkaloids, harmaline and harmine. 

Harmaline (C 13 H 14 N 2 0) was discovered in 1837 by Gobel, and harmine (C 13 H 12 N 2 0) in 
1847 by Fritche. According to the latter, the total yield of alkaloid is 4 per cent., of 
which two-thirds is harmaline and one-third harmine. 

Harmaline crystallises in yellow rhombic octahedra, neutralises acids, and forms 

* Fluckiger, Year-booh of Pharmacy, 1871, p. 600. 
TRANS. ROY. SOC. EDIN., VOL. XLVIL, PART II. (NO. 11). 37 



246 DR JAMES A. GUNN ON 

with them salts which have a yellow colour and bitter taste. Harmaline hydrochloride 
(C 13 H U N 2 0, HC1, 2H 2 0) forms long, fine yellow needles, easily soluble in water and in 
alcohol. Dilute solutions of harmaline exhibit a green fluorescence.* 

The only observations hitherto made to determine experimentally the general action 
of harmaline were those of Tappeiner in ] 8 9 5 . t His experiments were concerned 
chiefly with symptoms produced in frogs and mammals. 

In frogs he found that harmaline produces paralysis of voluntary movement. 
Reflex excitability remains after loss of voluntary motion, and even until arrest of the 
heart and respiration ; after its complete paralysis, nerve and muscle are quite excitable. 
No convulsions were observed. 

In mammals he found that harmaline produces motor disturbances, convulsions 
followed by paralysis. Consciousness is retained during the convulsions, but the 
sensitiveness to pain is reduced throughout the poisoning. Respiration is accelerated, 
and the temperature slightly raised. He found the minimum lethal dose to be O'll 
gramme per kilogramme, death being due to the arrest of respiration, which occurs 
suddenly without any previous considerable reduction in frequency. 

From two kymograph experiments on rabbits, he concluded that the blood pressure 
is at first considerably raised, especially during the convulsions. This rise is due to 
stimulation of the vasomotor centre, since the heart's action did not appear to be 
increased. Subsequently the blood pressure falls, owing to paralysis of the vasomotor 
centre and increased weakness of the heart. 

He summed up as follows : " If we judge by the most apparent symptoms of 
poisoning — the convulsions — which seem to be direct, as they are due to disturbance 
neither of the respiration nor of the circulation, we can until further notice reckon 
harmaline and harmine as convulsive poisons, and, in so far as death is due to arrest of 
the respiration, also as respiratory poisons. This investigation gives no basis for its 
therapeutic uses." 

In 1899 Oscar RaabJ investigated the action on Infusoria of several substances 
which form fluorescent solutions, e.g. acridin, phosphine, quinine, and eosine. He 
found that all of these are much more toxic to Paramcecium caudatum in light than 
in darkness, those light rays which most excite fluorescence being especially powerful 
in increasing the toxicity. In 1903 he extended these investigations to include 
harmaline, § and found that a solution of 1 in 20,000 of harmaline kills paramcecia in 
8 to 20 minutes, the presence or absence of light having no effect on the toxicity. 
Nor does the light effect come into play with 1 in 40,000 ; but when exposed to a 
solution of harmaline hydrochloride of 1 in 200,000, paramcecia remain quite healthy 
in the dark for 20 hours, while in the light they die in 1 to 3 hours. Exposure to 
light has by itself no injurious action on paramcecia. 

* Husbmann, Die Pflanzenstoffe, 1871, p. 76. 

t Archivfur exper. Pathol, u. Pharmakologie, Bd. xxxv., 1895, p. 69. 

1 Zeitschrift fur Biologie, Bd. xxxix., 1899, p. 524. § Ibid., Bd. xliv., 1903, p. 16. 



THE PHARMACOLOGICAL ACTION OF HARMALINE. 



247 



In 1901 Jacobson* showed that, in regard to their toxic effect on ciliated 
epithelium, solutions of harmaline hydrochloride (among other fluorescent substances) 
act independently of light in stronger concentrations, while in weaker concentrations 
they are much more poisonous in light than in darkness. 

In August 1908 I obtained from Dr J. F. Thorpe, F.R.S., of Manchester University, 
7 grammes of harmaline and 5 grammes of harmine for pharmacological investigation, 
for which kindness I take this opportunity of expressing my great indebtedness to him. 

After performing a large number of experiments with harmaline, both with regard 
to its general effects and its effects on isolated tissues, I have come to the conclusion 
that the pharmacological actions of this alkaloid, almost without exception, resemble 
intimately those of quinine ; and it is possible that this more extended investigation 
may give, contrary to Tappeiner's conclusion, some pharmacological basis for the 
therapeutic use of harmaline. Hence, in the course of the following account of the 
pharmacology of harmaline, its various actions will be compared in detail with those of 
quinine. 

A. Lethality of Harmaline. 

The lethality of harmaline was determined for frogs, guinea-pigs, rabbits, rats, and 
cats, with the following results : — 

Table I. — Minimum Lethal Dose by Subcutaneous Injection for Frogs. 



No. of 


Weight of 


Dose per 


Actual Dose 






Experi- 


Frog 


Kilogramme 


in 


Result. 




ment. 


in Grammes. 


in Grammes. 


Grammes. 






1 


31 


0-05 


0-0015 


Recovery. 




2 


34 


01 


00034 






3 


20 


02 


0004 






4 


26 


02 


0-0052 






5 


38 


0-25 


0-0095 


Death in 31-36 hours. 




6 


32 


03 


0-0096 


„ before 20 l„ 




7 


28 


04 


0-011 


„ in 6 hours. 




8 


22 


045 


o-oi 







Table II. — Minimum Lethal Dose by Subcutaneous Injection for Guinea-pigs. 



No. of 
Experi- 
ment. 



9 

10 
11 
12 
13 



Weight of 

Guinea pig 

in Grammes. 



500 
450 
470 
500 
500 



Does per 
Kilogramme 
in Grammes. 



001 
004 
08 
0-1 
0-2 



Actual Dose 

in 

Grammes. 



0-005 

0018 

0-038 

0-05 

01 



Result. 



Recovery — slight effects. 

,, marked effects. 

„ severe effects. 

Death in 7 hours 40 minutes. 

„ 33 minutes. 



Zeitschrift fur Biologic, Bd. xli., 1901, p. 445. 



248 



DR JAMES A. GUNN ON 



Table III. — Minimum Lethal Dose by Subcutaneous Injection for Rabbits. 



No. of 


Weight of 


Dose per 


Actual Dose 




Experi- 


Rabbit 


Kilogramme 


in 


Result. 


ment. 


in Grammes. 


in Grammes. 


Grammes. 




14 


1120 


0-02 


0-022 


Recovery — slight effects. 


15 


1650 


0-08 


0132 


,, severe effects. 


16 


1150 


009 


0103 


>i )> >> 


17 


1000 


01 


01 


Death in 1 hour 30 minutes. 


18 


1450 


012 


0174 


„ 3 hours. 



Table IV. — Minimum Lethal Dose by Subcutaneous Injection for Rats. 



No. of 
Experi- 
ment. 


Weight of 

Rat 

in Grammes. 


Dose per 
Kilogramme 
in Grammes. 


Actual Dose 

in 

Grammes. 


Result. 


19 
20 
21 

22 


132 
150 
170 
107 


0-076 
0-09 

o-ii 

012 


o-oi 

0-0135 

0-019 
0013 


Recovery — severe effects. 
Death in 1\ hours. 



Table V. — Minimum Lethal Dose by Subcutaneous Injection for Cats. 



No. of 
Experi- 
ment. 


Weight of 

Cat 

in Grammes. 


Dose per 
Kilogramme 
in Grammes. 


Actual Dose 

in 

Grammes. 


Result. 


23 
24 
25 


2700 
3000 
1730 


0-09 

0-1 

013 


0-243 

0-3 

0-225 


Recovery— severe effects. 
Death in about 9 hours. 
„ \\ hours. 



For determination of the minimum lethal dose in frogs, and for all subsequent ex- 
periments on frogs, the species Rana tem'poraria was used. In frogs, injections were 
made into the dorsal lymph-sac, and in mammals under the skin of the right flank. 

Minimum Lethal Dose for Frogs.- — Recovery followed from doses of 0'2 gramme 
per kilogramme and under; doses of 0'25 gramme per kilogramme and above proved 
fatal. The minimum lethal dose is therefore about 0'25 gramme per kilogramme. 

Minimum Lethal Dose for Mammals. — In the cases of the guinea-pig, rabbit, and cat, 
doses of 0"09 gramme per kilogramme and under were followed by recovery ; doses of 0'1 
gramme per kilogramme and above proved fatal. The minimum lethal dose is there- 
fore 0"1 gramme per kilogramme. In the case of the rat, recovery followed from doses 
of 01 1 gramme per kilogramme and under, while 0"12 gramme per kilogramme was 
the smallest dose to prove fatal. In this mammal the minimum lethal dose is therefore 
slightly higher, namely 012 gramme per kilogramme. 



THE PHARMACOLOGICAL ACTION OF HARMALINE. 249 

Harmaline resembles quinine in being relatively much more toxic to mammals 
than to frogs. 

B, Symptoms produced by Harmaline. 

(a) In Frogs. 

The following experiment will serve to illustrate the symptoms produced by a 
minimum-lethal dose of Harmaline. 

Experiment 5. — Rana temporaria, male, weight 38 grammes. At 11.27 a.m. the 
throat respirations were 1 7 in ten seconds, deep and regular ; the cardiac impacts were 
7 in ten seconds, fairly easily seen. Faradic stimulation of the skin of the right leg 
with the secondary coil at 160 mm. elicited a slight kick of the same leg, and the 
animal moved away if the stimulation was prolonged ; similar stimulation with the 
coil at 140 mm. caused immediate extension of both legs. Stimulation of the skin over 
the dorsal part of the spinal cord caused extension of the legs with the coil at 140 mm. 

At 11.50, 0'0095 gramme of harmaline hydrochloride dissolved in 0*8 c.c. of 
Ringer's solution was injected into the dorsal lymph-sac. This was equivalent to 0'25 
gramme per kilogramme. 

At 11.58 the pupils, which had been medium before injection, were more con- 
tracted. The results of electrical stimulation were the same as before. There was 
persistent contraction of the muscles running from the skin to the side of the urostyle. 

At 12.5 p.m. the respirations were 13 in ten seconds, and very feeble. The back 
was rigid, and curved in a direction concave upwards, due to rigidity of the back 
muscles. The frog jumped away when the skin of the leg was stimulated at 230 mm., 
and turned immediately when placed on its back. 

At 12.14 the rate of the respirations had fallen to 9 in ten seconds. With the coil 
at 160 mm., stimulation of the skin of the right leg caused extension of both legs, 
while at 12.34 the coil required to be at 145 mm. to produce this effect. 

At 1.50 the pupils were contracted. An area of skin overlying the dorsal lymph- 
sac was pale, contrasting markedly with the much darker colour of the skin of the 
rest of the body. The respirations were only 3 in ten seconds, and consisted of very 
feeble undulations of the anterior part of the floor of the mouth. The cardiac impacts 
were 3 in ten seconds. The reflex excitability as determined by electrical stimulation 
was unchanged. When the frog was laid on its back it made no effort to recover, but 
when the foot was now stimulated the frog turned over with difficulty. The animal 
jumped only if strongly stimulated, and when it alighted the hind limbs were not 
drawn up with normal rapidity. 

At 2.25 the respirations had ceased, and the cardiac impacts were not visible, but 
when the web of the foot was examined under the microscope the circulation was seen 
to be active. At 3.15 the animal lay prone, with limbs extended. The conjunctival 
and nose reflexes were active. 



250 DR JAMES A. GUNN ON 

At 6.30 p.m. no reaction resulted from either touching the conjunctiva or stroking 
the nose. The eyelids were fully open, and the pupils semi-dilated. 

At tl a.m. next day, as also at 12.30 p.m., with the coil at 100 mm., stimulation 
of the skin of either leg caused feeble movements of both legs, and stimulation of the 
skin over the dorsal part of the cord caused fairly vigorous extension of both legs. 
The circulation in the web of the foot was fairly active. 

At 3.20 p.m. stimulation of the skin of either leg with the coil even at 40 mm. 
produced no movements of the opposite leg. Stimulation over the cord at 100 mm. 
still produced feeble extension of both legs. 

The heart was now exposed by removing part of the sternum, and found to be 
beating moderately strongly at the rate of 2 in ten seconds ; auricles and ventricle were 
beating synchronously. 

The left sciatic nerve was exposed, and stimulation of it with the coil at 295 mm. 
induced contraction of the left gastrocnemius muscle, but no movement of the opposite 
limb ensued even with the coil at 40 mm. The muscles reacted to direct stimulation 
at 50 mm. 

At 5.0 p.m. the heart was beating feebly at the rate of about 4 in 30 seconds. No 
crossed reflex movements could be obtained by stimulation either of the skin of the leg 
or the sciatic nerve. Stimulation over the cord with the coil at 80 mm. caused feeble 
localised movements of the leg muscles, and stimulation of the sciatic nerve at 130 mm. 
caused a feeble contraction of the gastrocnemius. 

At 10.0 p.m. the heart was found to be arrested in diastole, and the muscles and 
nerves were irresponsive to electrical stimulation. 

(b) In Mammals. 

There is a close similarity in the effects produced by harmaline in different 
mammals. To illustrate the general nature of these effects, the following two 
experiments were selected ; the former as an example of the symptoms produced by a 
minimum lethal dose, in the case of which death is postponed for several hours, the 
latter as an example of a rapidly fatal dose. They also illustrate differences in the 
symptoms in different animals. 

1. Guiriea-jngs. — Experiment 12. — Guinea-pig, male, weight 500 grammes. At 
11.5 a.m. the cardiac impacts were 37, the respirations 21, in ten seconds. 

At 11.45, 0'05 gramme of harmaline hydrochloride dissolved in 4 c.c. Ringer's 
solution was injected under the skin of the right flank. This was equivalent to 
01 gramme per kilogramme. 

At 11.50 the cardiac impacts were 30, the respirations 18, in ten seconds. There 
were slight tremors of the head, and the right hind limb was slipping on the tray. 
During the next five minutes the animal made several short, jerky rushes forward ; in 
the quiet intervals the hind limbs were extended as if unable to support the body- 
weight. There were marked tremors of the head, and chewing movements of the jaws. 



THE PHARMACOLOGICAL ACTION OF HARMALINE. 251 

At 12.0 the cardiac impacts were 21, the respirations 18, in ten seconds. During 
the next fifteen minutes the animal remained usually with the fore part of the body 
upright, but with the hind-quarters lying on the side, there being frequent clonic 
movements of the hind legs. When it was disturbed it could run normally, and when 
laid on its back it turned over immediately. It uttered cries frequently. 

At 12.20 the hind limbs were somewhat rigid, and in the extended position. 
Respirations seemed to be more panting, but their rate could not be counted owing to 
the incessant tremors and clonic movements. 

At 12.32 the conjunctival reflex was duller, and the convulsive movements less 
violent. The animal generally lay on its side, occasionally assuming the upright 
position for short periods. Two minutes later it lay on its back, and from this time 
onwards usually remained lying on its back or side. 

At 1.0 the cardiac impacts were 20 in ten seconds. There were frequent clonic 
movements of the limbs and jaws. The tongue was often quickly protruded and with- 
drawn. There were sometimes rapid movements of the eyelids and oscillations of the 
eyeballs. When the animal was placed in the upright position it made no effort to 
move forward but slowly fell over on to its side. There was apparently a definite 
diminution in reflex excitability, for even pricking the skin of the leg with a pin 
induced no reaction. 

At 3.0 the limb movements were much feebler, though still almost constant. The 
conjunctival reflex was dull. The body was felt to be abnormally cold. The respira- 
tions were about 20, and the cardiac impacts about 18, in ten seconds, but fchey were 
difficult to count accurately. At 4.25 the rectal temperature was 26° C. 

At 7.13 the animal was lying on its side gasping. The respirations were 2 or 3 in 
ten seconds, irregular in rate and accompanied by gaping of the mouth. The cardiac 
impacts could not be felt. There were feeble pawing movements of the fore limbs. 
The hind limbs were quite stiff, as if in rigor mortis. The conjunctival reflex was 
completely gone, but the animal moved its head feebly when the ear was pinched. 
The rectal temperature was 21° C. 

At 7.25 all respiratory movements ceased. The thorax was then opened, and the 
heart was exposed at 7.29 and found to be beating slowly and feebly. 

2. Cats.— Experiment 25.— Cat, male, 1730 grammes. At 11.35 the respirations 
were 6, the cardiac impacts 34, in ten seconds. 

At 11.40, 0*225 gramme of harmaline hydrochloride dissolved in 2 c.c. Ringer's 
solution at about body temperature, was injected under the skin of the right flank. 
This was equal to - 13 gramme per kilogramme. 

At 11.43 the cat was timid and uneasy, and appeared to have hallucinations. He 
moved sometimes in a circle, trembling and with eyes staring. There were slight 
tremors of the body. 

At 11.48 he stood mewing loudly, with back arched and limbs extended and 
upright. The mouth was widely opened, and saliva flowed from it. 



252 DE JAMES A. GTJNN ON 

At 11.55 he had a sudden, violent convulsion which lasted about thirty seconds, 
and during which lie fell on his side. After this he remained lying on his side, mewing 
loudly. The respirations were then easily counted, and were 16 in ten seconds, deep 
and regular. 

Thereafter the animal had convulsions at 12.1, 12.2, 12.3, 12.5, 12.9, 12.13, 12.20, 
12.25, and 12.30. These convulsions conformed to a fairly definite type. They 
generally began by gaping of the mouth, followed suddenly by rapid clonic movements, 
involving especially the hind limbs ; the animal generally fell on its side. Often 
ushered in by a cry, with occasionally a short tonic phase prior to the clonic movements, 
and followed by an interval of flaccid ity during which the respirations were panting, 
these convulsions frequently bore a strong resemblance to attacks of grand mat. It 
is also seen that there is a tendency to gradual lengthening of the interval between 
the attacks. The later convulsions also became less violent, and the animal during the 
intervals more paretic. Up to 12.20 there was no impairment of the conjunctival reflex. 

At 12.25 the respirations were 20 in ten seconds, regular and deep. The con- 
junctival reflex was duller, and pinching the tail evoked no reaction. The fore limbs 
were extended at right angles to the body, while the hind-quarters lay on the side. 

At 12.30 the respirations were 14 in ten seconds, and much feebler. The cardiac 
impacts were 27 in ten seconds. 

At 12.44 no reaction resulted from strongly pinching the limbs or tail; the 
conjunctival reflex was almost abolished, but the eyes were closed if blown upon. 

At 1.3 the respirations were 15, the cardiac impacts 25, in ten seconds. The cat 
was lying quite quiet, and there were no tremors or convulsions. 

At 1.10, apart from the movements of respiration, the animal appeared completely 
paralysed. When held up by the neck it hung quite limp, and when placed on the 
floor made no attempt to support itself. 

At 1.13 the respirations were 4 in ten seconds, gasping in character; the cardiac 
impacts were 15 in ten seconds, feeble and irregular. At 1.14 respirations were 
reduced to an occasional gasp, and no cardiac impacts could be felt. 

At 1.16 the respirations ceased, and the pupils, which had been medium both 
before injection and throughout the poisoning, dilated widely. The thorax was now 
opened and the heart exposed ; there were no contractions of either auricle or ventricle. 
At 1.18, faradic stimulation of the phrenic nerve with the secondary coil at 550 mm. 
induced contraction of the diaphragm; at 1.22, stimulation of the sciatic nerve with 
the secondary coil at 370 mm. caused contraction of the gastrocnemius muscle. 

C. Action on the Cerebro-Spinal Nervous System. 

(a) Brain and Spinal Cord. 

I. In Frogs. — The frog is not a suitable animal for manifesting the action of drugs 
on the cerebrum, and no symptoms were observed which could be ascribed definitely 



THE PHARMACOLOGICAL ACTION OF HARMALINE. 253 

to an action of harmaline on this part of the brain. However, the onset of inco- 
ordination of movement, the loss of the power of jumping and of recovering the 
normal posture when the animal is laid on its back, and the cessation of respiration — 
occurring as these effects do at a time when the spinal reflexes and peripheral motor 
mechanism are slightly, if at all, impaired — indicate that harmaline first paralyses the 
functions of the mid brain and medulla oblongata. 

With regard to the spinal cord, a transient stnge of increased reflex excitability 
usually occurs, after which spinal reflexes become less and less easily elicitable. The 
voluntary muscles still readily respond to weak faradic stimulation of their nerves 
when the abolition of reflex excitability first occurs, showing that the paralysis is one 
involving the spinal cord. 

As is the case with many poisons, the power of transmission of impulses across the 
spinal cord, e.g. from left to right leg, is lost much earlier than the power of 
conducting impulses down the spinal cord, rendering it probable that the block in 
conduction is due not to paralysis of the efferent nerve cells but to interference between 
the afferent and efferent nerve cells of the cord. Loss of reflex excitability occurs 
before arrest of the heart, if the heart be not exposed. 

Harmaline resembles quinine very closely in its action on the central nervous 
system of the frog. Like harmaline, quinine in large doses paralyses the brain and 
respiratory centre, and later the reflex excitability of the spinal cord after a pre- 
liminary increase of excitability. In quinine poisoning too the heart continues to beat 
after paralysis of the spinal cord. 

2. In Mammals. — Harmaline affects the central nervous system of mammals in a 
manner different from the central nervous system of frogs, in so far as symptoms of 
excitation are added to symptoms of paralysis. Convulsions do not occur in frogs, but 
are the most conspicuous effects produced in mammals. 

In the cat these take the form of more or less violent convulsions of epileptiform 
character, occurring at irregular intervals and separated by quiescent periods. In the 
guinea-pig, clonic convulsions, less violent than those produced in the cat, and 
resembling running movements, occur almost without respite. In the rabbit, con- 
vulsive movements of an intermediate type are observed ; and in the rat, spasticity of 
the limb muscles, with tremors and swaying of the head and body, is the most 
characteristic appearance, though the kind of clonic convulsions described in the 
guinea-pig occurs also in the rat. 

Certain facts in regard to these convulsions aid in the localisation of the site of 

their production. In the first place, they are not due to asphyxia, because they occur 

before there is any impairment of respiration or any appearance of cyanosis. In the 

second place, the convulsions are cerebral in origin and not spinal. They are quite 

different from the convulsions produced by strychnine, for example ; they are not 

evoked by any apparent external stimulus, opisthotonus is never seen, and the 

convulsions do not occur in frogs. Also it will be shown in kymograph experiments 
TRANS, ROY. SOC. EDIN., VOL. XLVII. PART II. (NO. 11). 38 



254 DR JAMES A. GUNN ON 

that the convulsions do not occur in an animal anaesthetised by ether or chloroform. 
Apart from negative evidence, the epileptiform character of the convulsions points to 
the brain as the seat of their origin. It is, moreover, probable that they are due 
mainly to an exciting action on the cerebral cortex. The occurrence in cats of 
symptoms apparently due to hallucinations suggests an action on the cerebral cortex. 
Further, if the convulsions were due to an action on the lower part of the brain, they 
would probably occur in frogs, as in the case of the convulsions produced by 
picrotoxin. 

Tremors or convulsions are not produced in rabbits by doses less than 0"02 gramme 
per kilogramme, i.e. one-fifth of the minimum lethal dose ; in the guinea-pig, tremors 
occur after doses of above O'Ol gramme per kilogramme. 

In none of my experiments were the convulsions observed to be fatal, probably 
because their clonic nature does not seriously interfere with respiration. 

After a period, which varies with animal used and the dose given, the convulsions 
pass off, and in the case of smaller doses recovery ensues. In the case of larger doses 
the convulsions diminish in severity and are succeeded by a stage of exhaustion and, it 
may be, of complete motor paralysis, which may endure for some considerable time 
before failure of the heart or respiration. When this stage of motor paralysis super- 
venes, the symptoms referable to the central nervous system approximate closely to 
those described in frogs, and are probably due to like causes. There is possibly an 
increased excitability of the spinal reflexes in the early stages of poisoning in mammals 
as in frogs, but of this it is difficult to be certain. However, it is clear that the 
ultimate effect on the spinal cord, that of paralysis, is the same in mammals as in frogs. 
This is well seen in Experiment 25 described above, where the spinal reflexes were 
abolished some time before death, while the fact that the excitability of the peripheral 
neuromuscular mechanism was unimpaired even some minutes after death showed that 
this effect was central in origin. Owing to this paralysis of the spinal cord, the 
terminal phenomena of asphyxia are unattended by convulsions. 

Though convulsions, due to an action on the cerebrum, may occur under certain 
circumstances also with quinine, they are more typical of some of the other cinchona 
alkaloids, e.g. cinchonamin. Harmaline and quinine are, however, alike in producing, 
in large doses, diminution of excitability of the spinal cord in mammals. 



(b) Nerve. 

To investigate the action of harmaline on nerve structure, the following method 
was employed. The muscle response was taken as an index of the condition of the 
nerve. The gastrocnemius muscle of the frog was used, with the sciatic nerve carefully 
removed leaving a small part of the spinal column attached. Two pairs of electrodes, 
proximal and distal, were used to stimulate the nerve, and a small part of the nerve 
between the proximal and distal electrodes was painted with a solution of harmaline. 



THE PHARMACOLOGICAL ACTION OF HARMALINE. 255 

A comparison was made, before and after the application of harmaline, of the muscle 
response evoked by proximal and distal stimulations. 

It was found that nerve is very resistant to harmaline : even so strong a solution as 
2 per cent, of harmaline hydrochloride impairs the conductivity only after several 
hours' application. This feeble action on nerve presents a marked contrast to the 
unusually toxic effect which harmaline exerts on striped muscle. 

As in the case of the nerve trunks, so the motor nerve terminals are slightly if at all 
affected by harmaline. Thus in Experiment 5, in the course of impairment of the 
peripheral neuromuscular mechanism, stimulation of the sciatic nerve of the frog caused 
contraction of the gastrocnemius with the coil at 295 mm., while direct stimulation of 
the muscle required a strength of 50 mm. This relation of excitability appears to show 
that the peripheral paralysis, which occurs late in the course of poisoning, is due to 
paralysis of the muscle, the nerve and nerve-ends being slightly if at all affected. This 
opinion was confirmed by investigating the action of harmaline on the nerve-muscle by 
Claude Bernard's method, when it was found that diminution in excitability of muscle 
when stimulated through its nerve occurred only when there was a corresponding 
decrease in the excitability of the muscle to direct stimulation. 

D. Action on Skeletal Muscle. 

It has been shown that, after injection of harmaline, rigor and loss of excitability 
occur in those muscles to which the injected solution obtains more direct access. This 
produces an appearance of opisthotonos when the injection is made into the dorsal lymph- 
sac, of emprosthotonos when the injection is made under the skin of the abdomen. 
Arrangements were made to test further this action of harmaline by keeping a muscle 
directly in contact with solutions of varying strengths of the alkaloid. A modified 
Wild's method was employed ; and when the muscles were stimulated, the current from 
the secondary coil passed simultaneously through both muscles. Tracings were taken 
on a slowly revolving drum. 

Experiment 26 (figs. 1 and 2). — Strength of solution, 1 in 500. Normal twitches 
resulting from stimulation of the muscles with break shocks are shown at 11.40 (fig. 1). 
At 11.45, Ringer's solution was withdrawn from the upper tube, and a solution of 
harmaline hydrochloride (1 in 500 of Ringer's solution) was substituted. Almost 
immediately this caused the muscle to pass into rigor. This proceeded so rapidly that 
in five minutes this muscle had raised the lever above the level of the summit of 
a single twitch. 

At 11.58, thirteen minutes after exposure to harmaline, the upper muscle ceased to 
respond to stimulation, whereas the control muscle contracted as before (fig. 2). 

This strength of solution therefore very soon brings on rigor with loss of 
excitability of the muscle. 

Rigor of frog's muscle is produced by solutions of harmaline even so dilute as 1 in 



256 



DR JAMES A. GUNN ON 



10,000, or sometimes 1 in 20,000. Harmaline produces rigor as rapidly and as com- 
pletely in curarised muscle, so that the effect is probably due to an action on the muscle 
protoplasm. In respect of this action in producing shortening and loss of excitability 
of muscle, harmaline resembles quinine, but is more powerful. In none of the experi- 
ments did fibrillary twitchings of the poisoned muscle occur. 

Solutions of harmaline which are too dilute to produce rigor of muscle do not 




Fig. 1. 



Fig. 2. 



markedly diminish the excitability of the muscle. Thus a gastrocnemius-sciatic 
preparation was kept in a solution of harmaline 1 in 25,000 for eighteen hours, at the 
end of which time the muscle reacted to direct and indirect stimulation almost as well 
as a control preparation kept for the same time in Ringer's solution alone. This may 
explain why, in spite of the fact that actual rigor of muscle may be produced by a 
solution so dilute as 1 in 20,000, paralysis of the voluntary muscles plays a part 
relatively so unimportant among the general effects of harmaline poisoning, for the 
muscles respond to stimulation of their nerves for some hours after abolition of reflex 
excitability. 

E. Action on the Circulation. 

(a) Heart. 

The description given of the symptoms occurring in the frog after subcutaneous injec- 
tion of harmaline has shown that this alkaloid at an early period causes slowing of the 
rate of the heart, but that the heart, though apparently much enfeebled, continues to beat 
until after cessation of respiration and paralysis of reflex excitability. In mammals too 
the rate of the heart-beats is reduced very soon after the injection of harmaline. With 
large doses, this reduction in rate is progressive, and arrest of the heart plays a part 



THE PHARMACOLOGICAL ACTION OF HARMALINE. 



257 



almost as important as arrest of the respiration in causing death. In the case of 
smaller lethal doses, where death is clearly due to respiratory failure, the heart rate is 
diminished soon after injection to a certain extent, and tends to remain at this point 
till the respirations fail. For example, in Experiment 1 8, where a rabbit received a dose 
which killed in three hours, the cardiac impacts were reduced from 46 to 27 per ten 
seconds in twenty minutes, and were still 27 per ten seconds two hours later. The 
heart is always arrested in diastole. In several experiments in which the heart 
continued to beat after arrest of the respirations, the right vagus was exposed, and 
stimulation of it with the coil at 70 mm. to 90 mm. was found to arrest the heart-beats. 

Further experiments were made to ascertain the action of harmaline on the isolated 
heart, the frog's heart being used. 

In a first series of experiments the heart was perfused in situ through the hepatic 
vein, the perfused fluid escaping through the cut aorta ; and in a second series the 




*'<*t *77me = Mint/tcS 



A-27 



Fig. 3. 

isolated ventricle alone was perfused by means of Schafer's frog-heart plethysmograph. 
Only the latter series need be described, as the results were similar in both. As the 
nutrient solution and as the solvent for harmaline a mixture of defibrinated ox blood 
(one part) and Ringer's solution (two parts) was used. The bulb of the plethysmograph 
which contained the heart was filled with Ringer's solution, and the contractions of the 
ventricle were recorded by means of an air-piston recorder attached by a rubber tube to 
the brass cylinder. 

Experiment 27 (fig. 3). — Strength of solution, 1 in 2000. 

A solution of 1 in 2000 within one minute arrested the contractions of the ventricle 
in almost complete systole (fig. 3). Two minutes later the normal solution was turned 
on, and the heart slowly dilated till it reached the condition of relaxation normal to the 
end of diastole, whereupon it again resumed beating. The contractions gradually 
improved till in fifteen minutes the rate and excursus were practically the same as 
before harmaline. A second introduction of the harmaline solution produced the same 
effect as before. 



258 



DR JAMES A. GUNN ON 



Experiment 28 (tigs. 4 to 7 inclusive). — Strength of solution, 1 in 10,000. 

This solution very soon reduced the rate of beat of the ventricle, and to a less extent 




Fig. 4. 



the amplitude of its excursus (figs. 4, 5, and 6). In twenty-two minutes the ventricle 
ceased contracting altogether, in a position of almost complete relaxation. Two minutes 




Fig. 5. 



Fig. 6. 



later the normal solution was turned on, and in five minutes this completely reinstated 
the ventricle to its normal action (fig. 7). A second introduction of the harmaline 




Fig. 7. 



solution produced the same poisoning effects more quickly, and the ventricle again 
completely recovered when this was replaced by the normal solution. The normal 
solution was now perfused through the ventricle for over two hours. To the harmaline 
solution was then added atropine sulphate in the proportion of 1 in 100,000. This 



THE PHARMACOLOGICAL ACTION OF HARMALINE. 



259 



combined solution brought about arrest of the ventricle in the same way as was pro- 
duced previously by the same strength of harmaline without atropine. 



These effects are detailed in the following table :- 



Table VI. — Experiment 28. 





Rate 


Amplitude 






Time. 


per 
Minute. 


of 
Excursus. 


Notes. 




12.33 


22 


25 






12.36.30" 






Solution of H.H. 1 in 10,000 turned on (fig. 4). 




12.40 


16 


24 






12.48 


13 


19 






12.53 


8 


19 


Fig. 6. 




12.59 












1.1 






Normal solution turned on (fig. 7). 




1.2 


11 


15 






1.3 


19 


14 






1.7 


23 


25 






1.14 


27 


25 






1.15 






Solution of H.H. 1 in 10,000 turned on. 




1.17 


7 


25 






1.20 


2 


23 






1.22 


1 


20 






1.23 












1.25 




• • • 


Normal solution turned on. 




1.26 


11 


18 






1.31 


26 


25 






3.48 


27 


24 






3.51 






Solution containing H.H. 1 in 10,000 and A.S. 1 
turned on. 


in 100,000 


3.53 


2 


20 






3.54 













Experiment 29. — Strength of solution, 1 in 20,000. 

This solution, in three and a half hours, reduced the rate of beat from 25 to 9 per 
minute, and the amplitude of excursus from 28 mm. to 21 mm. Complete recovery was 
then brought about in fifteen minutes by re-introduction of the normal solution. The 
harmaline solution, to which had been added atropine sulphate in the proportion of 
1 in 50,000, was then perfused, but the addition of atropine did not prevent the slowing 
of the ventricle produced by harmaline. 

Experiment 30. — Strength of solution, 1 in 50,000. 

Perfusion of the ventricle with this solution had practically no effect on the heart 
in two and a half hours. 

These experiments show that very strong solutions of harmaline bring about 
abrupt systolic arrest of the heart, an effect probably similar in nature to the rigor of 
voluntary muscle which strong solutions of harmaline have been shown also to effect. 

The characteristic action of harmaline on the heart, however, is to slow the rate 



260 



DR JAMES A. GUNN ON 



and diminish the completeness of systolic contraction, actual arrest in almost complete 
diastole being produced by a solution of 1 in 10,000. 

The slowing of the heart is uninfluenced by simultaneous perfusion with atropine, 
and is therefore probably due to an action on the cardiac muscle, and not to stimula- 
tion of the vagal terminations. 

In its action on the frog's heart, harmaline resembles quinine, which produces, in 
much the same concentrations, slowing of the heart (which is not prevented by 
atropine) and arrest in diastole. 

(b) Blood-vessels. 

To ascertain any changes produced by harmaline in the blood-vessels of the frog, 
the following method was used. After exposure of the heart of a pithed frog, the venae 



1 










































































































H 


u 








































































































C- 


C. 




















































































































































































































i 


>— i 


►— « 


i— i 














s 


s 




















» , 








































































































f 




































ft 








f 




>* 




r 












5 


















































































l 9 






















































































( 


y-i 


X 


i>-l 


m: 


> -■ 




N 


^ 


►-7 


?u 


i£ 


isr 


£ 


»>cC<<iiin, 


' 
























































































































































































































r\ 








































































































-i 


V 
















































































































































































































































































































































































































































































































































t 
























































































































































































































































































































































































































































































































































m 


lyi 


XX. 


Xi 


■o 


— 






/ 


•? 


















X 


o 


















3 


O 


















4 


o 


















6~ 


o 



Fig. 8. 

cavae were cut across and a fine cannula was tied into the left aorta, the right aorta 
being ligatured. This cannula was connected with a series of Mariotte's flasks which 
contained the fluids to be perfused. The amount of fluid exuding per minute from 
the cut venae cavae was accurately measured. Ringer's solution was used as the 
normal solution and as the solvent for harmaline. 

The results of these perfusion experiments may be shortly stated as follows : — 
A solution of 1 in 1000 reduced the flow from 2 '6 c.c. per minute to 0*6 c.c. per 
minute in thirty minutes ; a solution of 1 in 5000 (see fig. 8) reduced the flow from 
3 "2 c.c. per minute to 0*9 c.c. per minute in thirty minutes ; a solution of 1 in 10,000 
reduced the flow from 2'5 c.c. per minute to 17 c.c. per minute in forty minutes; 
a solution of 1 in 20,000 reduced the flow from 2*1 c.c. per minute to 1*5 c.c. per 
minute in sixty minutes ; while solutions of 1 in 25,000 and 1 in 40,000 had no effect 
on the vessels in forty minutes. 

As there occurred in the frogs during these perfusions practically no oedema, it is 



THE PHARMACOLOGICAL ACTION OF HARMALINE. 



261 



apparent that harmaline causes constriction of the arterioles : a marked constriction in 
the case of solutions not less dilute than 1 in 5000, and a still perceptible constriction 
with a solution of 1 in 20,000. I have recently shown * that quinine exerts a similar 
action on the blood-vessels of the frog. 

(c) Heart and Blood-vessels. [Blood Pressure.) 

In all blood-pressure experiments the animals (rabbits or cats) were first anaesthet- 
ised with chloroform ; the trachea was then exposed, and a cannula tied into it through 
which diluted ether was thereafter inhaled. A cannula in the left carotid artery was 
connected ||with the manometer. Respirations were recorded by means of a double 




Hdi-tnaSM^tj HyThxUUa-r^f 



°-°S. frr. j=f-r R^, 



Fig. 9. 

stethograph attached by a band round the thorax and connected with a Marey's tambour. 
Iujections were made into the right jugular vein. 

Experiment 31 (fig. 9). — Rabbit, 2200 grammes. 

At 12.35, 0'044 gramme of harmaline hydrochloride dissolved in 2 c.c. Ringer's 
solution was injected. This was equal to 0'02 gramme per kilogramme. 

This experiment illustrates the effect of a rapidly lethal dose. Blood pressure, 
after a slight transient rise, rapidly falls. The respirations and heart-beats quickly 
decrease in rate. The pulsations in the carotid diminish in size in spite of the slowing 
of the heart. Death is due both to cardiac and respiratory failure. 

Experiment 32 (Table VII., figs. 10 to 14 inclusive). — Rabbit, 2000 grammes. 

Successive actual doses of 0"002 grm., 0'004 grm., 0*008 grm., and 0'016 grm., each 
dose being dissolved in 2 c.c. of Ringer's solution. This experiment shows the effects 
of increasing sub-lethal doses on the blood pressure and the respirations of the 
rabbit. 

* Archives internal, de Pharmacodynamie, 1909, p. 319. 
TRANS. ROY. SOC. EDIN., VOL. XLVII. PART II. (NO. 11). 39 



262 



DR JAMES A. GUNN ON 



Doses of 0*001 grm. per kilo and 0"002 grm. per kilo (fig. 10) caused a rise of blood 
pressure. Doses of 0'004 grm. per kilo and 0*008 grm. per kilo (figs. 11 to 14) caused 
a preliminary fall, followed by a rise, with later again a fall of blood pressure. 

The rise of blood pressure is in the case of all doses accompanied by slowing of 
the heart. The increase of blood pressure must therefore be due either to increased 

Table VII. — Experiment 32. 






Time. 


Dose 
of 


Average 


Pulse 


Rate 
of 


Respiration 


Notes. 




Harmaline. 


B.P. in mm. 


Rate. 


Respirations. 


Excursus. 




1.0 




104 


40 


12 


3 mm. 




1.0.50" 




102 


41 


13 


3 „ 




l.l 


001 grm. per kilo 












1.1.20" 




106 


35 


11 


3 mm. 




1.2 




105 


29 


11 


3 „ 




1.3 




102 


29 


10 


3 „ 




1.9 




107 


31 


11 


3 „ 




1.11 




106 


32 


10 


2 




1.12 


002 grm. per kilo 










Fig. 10. 


1.13 


... 


lib 


27 


9 


2 mm. 




1.14 




110 


24 


9 


2 „ 




1.16 


... 


110 


26 


9 


2 „ 




1.21 




114 


26 


8 


91 

*~2 » 




1.25 


... 


114 


30 


8 


2i ,, 




1.29 


... 


112 


28 


7 


2A 




1.30 


0004 grm. per kilo 












1.30.30" 


... 


108 


26 


7 


2 mm. 




1.31 




116 


24 


7 


2 „ 




1.33 




112 


24 


6 


91 

z 2 >» 




1.35 




106 


26 


7 


2 „ 




1.39 




105 


26 


7 


9 




1.42 


... 


105 


26 


7 







1.44 




106 


25 


7 


2 „ 




145 


008 grm. per kilo 










Fig. 11. 


1.45.30" 


... 


76 


20 


4 


1 mm. 




1.46 




76 


18 


6 


1 „ 




1.48 




92 


19 


7 


1 

2 " 


Fig. 12. 


1.50 30" 




120 


20 


5 


i ,, 




1.52 




114 


20 


5 


i „ 




1.56 




100 


21 


7 


n ,, 




1.59 




101 


20 


6 


9 




2.6 




108 


21 


7 


3 „ 




2.8.10" 




109 


20 


7 


25 „ n 


Anaesthesia incomplete. 


2.9.40" 




119 


20 


7 


7 ., ?J 


Fig. 13. 


2.22 




102 


18 


6 


2 


Fig. 14. 

Experiment discontinued. 



cardiac output per single beat or to contraction of the blood-vessels. Perfusions of 
the frog's heart give no indication of augmented action of the heart, whereas a solution 
of harmaline so dilute as 1 in 20,000 constricts the frog's blood-vessels. It is probable, 
therefore, that the rise of blood pressure is due to contraction of the blood-vessels. 
The fact that this rise of pressure tends to be gradual and persistent points to the 
contraction of the blood-vessels being due rather to an action on the arterioles them- 



THE PHARMACOLOGICAL ACTION OF HARMALINE. 



263 



selves, as seen in the case of the frog, than to a stimulation of the vasomotor 
centre. 

In this experiment a total amount of 0'015 grm. per kilo was injected in divided 



FKWwnmwwrirtwwwwwWwVWVMl/y^^ 



ffffi^t^mimmmmiMwmmMtmftmtmwmm " 



VAAAAAA/WWVWWW\MM/WWWWWWVWW 



HA+irvUnS H&. O-OOZft*,. fa fCcU 



Fig. 10. 



doses — an amount which would immediately have been fatal if administered 
by a single injection. About forty minutes after the last injection, however, the 
blood pressure was still at the level normal to the animal, though the pulse rate had 




Fig. 11. 



diminished so greatly as from 42 to 18 per ten seconds. It would appear, therefore, 
that in the rabbit very large doses of harmaline may bring about a constriction of 
the arterioles so physiologically balanced that it may exactly maintain the level of 
blood pressure in spite of a profound fall of the pulse rate, 



264 



DR JAMES A. GUNN ON 



An important point, illustrated by this experiment and confirmed by all of a large 
number of blood-pressure experiments in rabbits and cats, is the complete absence of 







Fig. 12. 



convulsions in a properly anaesthetised animal. This was the only one of these 
experiments in which convulsions were ever observed ; they lasted for about one 



3.V. 



uUiJBmmiLki 



Resk 



•Miraiiwii uiiiiii 



w\\ 



moHmmmmm 




Fig. 13. 



Fig. 14. 



minute (see fig. 13), and were due to insufficient depth of anaesthesia. Tappeiner, in 
his two experiments on blood pressure, observed convulsions during practically the 
whole time of these experiments. Unfortunately he does not state what anaesthetic, 
if any, he used. This occurrence of convulsions vitiated the value of his experiments 



THE PHARMACOLOGICAL ACTION OF HARM ALINE. 



265 



as a study of the uncomplicated action of harmaline on the blood pressure and 
respiration. 

Experiment 33. — (Table VIII., fig. 15). — Cat, 3000 grammes. This experiment 
illustrates certain differences between the blood-pressure effects in the cat and those in 
the rabbit ; it also shows the terminal phenomena in a case where respiratory failure 
definitely precedes arrest of the heart (contrast fig. 9). 

At first, three injections were given, each of 0001 grm. per kilo. The first and third 
injections caused a slight rise of blood pressure, the second a fall. The effects of small 



Table VIII. — Experiment 33. 



Time. 


Dose 

of 

Harmaline. 


Average 


Pulse 


Rate of 


Respiration 


Notes. 


B.P. in mm. 


Rate. 


Respirations. 


Excursus. 


5.3 




164 


27 


9 


5 mm. 




5.3.40" 


0'001 grm. per kilo 












5.4 




175 


24 


9 


5 mm. 




5.5 




166 


24 


7 


4 „ 




5.6 




166 


25 


8 


4 „ 




5.7.30" 




170 


24 


8 


4 „ 




5.8 


001 grm. per kilo 












5.8.10" 




154 


22 


8 


4 mm. 




5.10 




158 


24 


8 






5.10.50" 


0"001 grm. per kilo 








4 mm. 




5.11 




162 


25 


9 


4 „ 




5.12 




158 


22 


8 


4J „ 




5.21 




150 


24 


8 


4 „ 




5.21.30" 


004 grm. per kilo 












5.21.50" 




138 


20 


7 


4 mm. 




5.23 




146 


20 


6 


4 ., 




5.29 




140 


24 


6 


4 „ 




5.37 




116 


30 


7 


4 „ 




5.40 




122 


27 


6 


4 „ 




5.40.30" 


01 grm. per kilo 










Fig. 15. 


5.50.30" 




96 


20 


9 


2 mm. 




5.51 




64 


19 


4 


2 „ 




5.52 




50 


18 


1 


1 „ 




5.53 




36 


16 





o „ 




5.57 













o „ 





doses are therefore inconstant, and it was found in other experiments that, in the cat in 
contrast to the rabbit, small doses of harmaline produce either a very slight rise, or 
sometimes no rise, of blood pressure. 

When this experiment is compared with Experiment 32, it is also seen that in the 
cat, after a total dose of 0'007 grm. per kilo, there is a considerable and permanent fall 
of blood pressure, whereas in the rabbit after the same dose blood pressure, after a 
temporary fall, recovers to the normal level. These differences are probably due to the 
fact that changes in blood pressure which are due to an action on the arterioles are more 
marked in the rabbit than in the cat, owing to the much greater length of the intestinal 



266 



DR JAMES A. GUNN ON 



canal in the former animal.* In the cat the constriction of the arterioles is insufficient 
to counterbalance the slowing of the heart, and the blood pressure falls. 



B/P 

mwmmmmmm 



AAAAAr>AAAA^AAAAAAAnAA/VAnyv^AAAAAAA/VVV> 



&J*SOLSS4. 



001 p*. j^ ffl( j o 




Fig. 15. 



With these experiments as examples of the effects of small and large doses, an 
account may now be given of further experiments made to define more accurately 
the nature of the production of these effects. 



The Cause of the Sloiving of the Heart. 

Experiment 34. — In this experiment, doses were given corresponding to those given 
in Experiment 32, with this difference, that sufficient quantities of atropine sulphate 
were administered during the experiment to ensure that, before each injection of 
harmaline, the terminations of the vagus nerves in the heart were paralysed. The 
conditions of the experiment are indicated in Table IX. 

In this experiment the first injection of O'OOl grm. per kilo reduced the pulse rate 
from 36 to 31 per ten seconds as compared with a reduction of from 41 to 29 per ten 
seconds by the same dose in Experiment 32, where no atropine was given. It seems 
probable that, in case of the first injection of a small dose of harmaline, part of the 
slowing of the heart is due to reflex stimulation of the vagus by the rise of blood 
pressure. 

On the other hand, doses of 0'002 grm. and 0'004 grm. per kilo produce an amount of 
slowing of the heart (when the terminations of the vagus are paralysed by atropine) 
quite commensurate with that produced by the same doses in Experiment 32, where the 

* Laudeh Brunton, Textbook of Pharmacology and Therapeutics, 1887, p. 288. 



THE PHARMACOLOGICAL ACTION OF HARM A LINE. 



267 



vagal mechanism was intact. The last dose (0'008 grm. per kilo) proved fatal in this 
experiment probably because respiratory failure was caused by the added effects of 
harmaline and atropine on the respiratory centre. However, a sufficient comparison can 



Table IX. — Experiment 34. 



Time. 


Injection. 


Average 
B.P. 


Pulse 
Rate. 


Result of Stimulation 
of Left Vagus. 


1.15 




100 


40 




1.16 








130 mm. — arrest of heart. 
140 „ slowing of heart. 
150 „ nil. 


1.18 


A.S. 001 grm. per kilo 








1.20 








140 mm. — nil. 
120 „ 


1.39 








140 ,, arrest of heart. 


1.40 


A.S. 0-002 grm per kilo 








1.40.30" 








140 mm. — nil. 
100 „ 


1.42 




94 


36 




1.43 


H.H. 0-001 grm. per kilo 








1.43.30" 




102 


33 




1.44.20" 


... 






120 mm. — nil. 


1.46 




102 


31 




2.15 








120 mm. — slowing of heart. 


2.25 








120 „ 


2.26 


A.S. 0-002 grm. per kilo 








2.29 








120 mm. — nil. 


2.29.30" 




79 


28 




2.30.10" 


H.H. 0002 grm. per kilo 








2.31 


... 


82 


24 




2.33 


. . . 


90 


23 


1 


2.36 




92 


22 




2.37 








120 mm. — nil. 


2.38 


H.H. 004 grm. per kilo 








2.38.30" 




84 


19 




2.39 




88 


20 




2.42 




96 


19 




2.45 








120 mm. — slowing of heart. 


2.52 


A.S. 0001 grm. per kilo 








2.55 




88 


22 




2.55.40" 


... 






120 mm. — nil. 


2.56 


H.H. 0-008 grm. per kilo 








2.56.40" 




52 


18 




2.58 




30 


13 




3.0 




22 


12 




3.3 













be made to allow the conclusion that, except in the case of the first injection, the slowing 
of the heart is not due to central or peripheral stimulation of the vagus, but to an 
action on the cardiac muscle. This result corresponds with that obtained in perfusion 
experiments on the frog's heart where the slowing of the heart produced by harmaline 
was not prevented by simultaneous perfusion with atropine. 



268 



DR JAMES A. GUNN ON 



The Cause of the Fall of Blood Pressure. 

It has been already stated that the rise in blood pressure produced by smaller doses 
of harmaline is to be attributed to contraction of the arterioles ; it remained to be 
determined whether the fall of blood pressure, when it occurs, is due to cardiac causes 
or to dilatation of the arterioles. 

Experiment 35 (fig. 16). — Rabbit, 1900 grammes. In this experiment a record 
was taken of the blood pressure and of the changes in intestinal volume. About four 
inches of the rabbit's small intestine was enclosed in an oncometer, and the changes 
in its volume were recorded by an air-piston recorder. A dose of 0"008 grm. per kilo 
was injected, and in thirty seconds this lowered the blood pressure from 100 mm. to 
78 mm., and the pulse rate from 19 to 16, per ten seconds. No increase of intestinal 




Fig. 16. 

volume occurred during this fall of pressure, which must therefore be ascribed to 
slowing or to slowing and weakening of the heart. 

Experiment 36. — Cat, 3300 grammes. A record was taken of the blood pressure 
and volume of the left kidney. A dose of 0'005 grm. per kilo was given, and this 
reduced the blood pressure in one minute from 109 mm. to 78 mm., and the pulse rate 
from 21 to 15 per ten seconds. During this fall of blood pressure there occurred 
a distinct diminution of kidney volume, showing that there was certainly no dilatation 
of the kidney vessels. In the cat, therefore, as in the rabbit, it is probable that the fall 
of blood pressure produced by large doses of harmaline is due to slowing or to slowing 
and weakening of the heart's action. 

The effects produced by harmaline on the blood pressure differ in some respects from 
those produced by quinine. With quinine "the heart is often accelerated at first, but 
is afterwards slow and weak, while the blood pressure, after a slight increase, declines 
progressively. The changes are caused by a preliminary contraction of the arterioles 



THE PHARMACOLOGICAL ACTION OF HARMALINE. 269 

and acceleration of the heart, followed by dilatation of the former and slowing and 
weakening of the latter."* Harmaline differs from quinine in its blood-pressure effects 
in that the preliminary rise of blood pressure is due only to contraction of the arterioles 
and not to acceleration of the heart, while the fall of blood pressure produced by large 
doses of harmaline is due, so far as my experiments have shown, only to cardiac 
slowing and weakening, and not to dilatation of the arterioles. 

F. Action on Respiration. 

Lethal doses of harmaline paralyse respiration both in frogs and in mammals. 
Since at the time of death in mammals faradic stimulation of the phrenic nerve causes 
tetanus of the diaphragm with a normally minimum stimulus (see Experiment 25), it is 
probable that respiratory failure is due to paralysis of the respiratory centre. 

Especially during the time that convulsions occur, the rate and vigour of the 
respirations are often increased. In none of my blood-pressure experiments (when 
convulsions were prevented by anaesthesia) was an increase in the rate or amplitude of 
the respirations observed, such as was found by Tappeiner. 

In the case of slowly fatal doses, death is due to arrest of the respiration alone, and 
the heart may continue beating as long as ten minutes after respiration has ceased. 
With larger doses, however, the heart-beats and respirations fail about the same time ; 
indeed, one or two respiratory gasps may occur after it is impossible to feel any cardiac 
impacts. 

Tappeiner stated that arrest of the respiration comes on somewhat suddenly, not 
being intimated by any previous considerable reduction in frequency. This is only 
partly true in the case of rapidly fatal doses, where the toxic effect on the heart attains 
greater prominence. In the case of smaller lethal doses, where heart failure does not so 
materially contribute to the cause of death, there is a gradual and progressive 
diminution in the rate of the respirations. 

G. Action on Temperature. 

Tappeiner stated that "the body temperature is rather heightened than lowered. 
In one case in a rabbit it was 40'4° C. and in a dog 397° C. in the rectum." It is 
unfortunate that he does not state with what doses or at what times after injection this 
elevation of temperature occurred, and also what was the temperature before injection. 
In the rabbit the normal temperature may be above 40 '4° C.t 

In my experiments I have not observed with any dose a definite rise of temperature, 
while large doses invariably produce a fall of temperature, in regard to which action 
harmaline resembles quinine. 

* Cushny, Textbook of Pharmacology, 3rd edit., p. 360. 

t Simpson and Galbkaith, Journal of Physiology, 1905, p. 230. 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART II. (NO. 11). 40 



270 



BR JAMES A. GUNN ON 



In the case of sub-lethal doses the fall of temperature is comparatively slight, e.g. 
the temperature of a rabbit which received four-fifths of the minimum lethal dose fell 
from 39° C. to 37° C. in 1^ hours. The fall is greater in the case of lethal doses ; e.g., 
the temperature of a rabbit which received 1^ minimum lethal dose fell from 38° C. to 
34° C. in one hour. However, the extent of the fall of temperature is not proportional 
to the dose in the case of lethal doses. The fall is progressive, and is the greater 
the longer the animal lives after injection. Thus, as was seen in the account of 
Experiment 12, the temperature of a guinea-pig which received a minimum lethal dose 




Fig. 17. 

was as low as 21° C. before death, which occurred 7§ hours after injection. Simpson 
and Herring * have shown that with such a low temperature a warm-blooded animal 
is narcotised by cold, and that when the temperature falls below 24° C. the animal 
cannot recover unless artificially warmed. Therefore in this case the profound fall of 
temperature is probably a contributory cause of death. 



H. Action on the Uterus. 



In view of the employment of harmaline in India to procure abortion, it was 
interesting to ascertain whether experimental evidence could be found of an action on 
the uterine muscle. In some experiments made for this purpose the following method 
was used. 

* Journal of Physiology, 1905, p. 305. 



THE PHARMACOLOGICAL ACTION OF HARMALINE. 271 

The animals (rabbits) were anaesthetised as for blood- pressure experiments, and 
kept during the experiment in a bath of saline solution at 39° C, just enough of the 
body being submerged to ensure that the uterus was never exposed to the air. The 
abdomen was then opened in the middle line, and the uterus, isolated from the 
surrounding viscera, was connected with a, lever writing on a slowly revolving drum. 

Experiment 37 (fig. 17). — Rabbit, 2100 grammes, non-pregnant. From 2.30 p.m. 
there were slight spontaneous contractions of the uterus, just sufficient to move the 
lever perceptibly. At 3.5 a dose of O'OOl gramme per kilogramme was given by the 
jugular vein (see fig. 17). This brought on almost immediately a very powerful tetanic 
contraction of the uterus lasting for about seven minutes, after which uterine con- 
tractions again remained in abeyance. At 3.25 a second injection was given of 0*0005 
gramme per kilogramme, which produced a similar tetanic contraction lasting for four 
minutes. 

This experiment shows that harmaline, in doses which previous experiments have 
shown to be the smallest to affect blood pressure, exerts a powerful action on the uterus 
in the direction of inducing vigorous and sustained tetanic contraction. This fact 
sufficiently explains the clinical observation that harmaline may cause abortion. It is 
an effect which is also, but less powerfully, produced by quinine. 

General Summary. 

The minimum lethal dose of harmaline hydrochloride by subcutaneous injection is 
for frogs 0*25 gramme per kilogramme, and for mammals (guinea-pig, rabbit, rat, and 
cat) about O'l gramme per kilogramme. 

In frogs, lethal doses of harmaline paralyse the mid-brain and medulla oblongata, 
and, at a much later period, the spinal cord. Abolition of reflex excitability occurs 
before arrest of the heart, and before paralysis of the voluntary muscles. 

In mammals, large doses of harmaline cause epileptiform convulsions varying 
somewhat in character in different animals. The convulsions are due to an action 
on the cerebrum, probably especially affecting the cortex. Lethal doses paralyse the 
spinal cord also in mammals, and, as in the case of frogs, this occurs at a later stage 
than the action on the brain. 

The conductivity of nerve is impaired only by prolonged direct application of 
strong solutions of harmaline, and the motor nerve-ends are slightly, if at all, affected 
by this alkaloid. On the other hand, harmaline in comparatively weak solutions 
causes rigor and excitability of voluntary muscle. 

When perfused through the frog's heart, strong solutions of harmaline cause almost 
immediate systolic arrest of the heart ; weaker solutions cause slowing of the heart and 
diminution of systolic contraction, and arrest of the heart, when it occurs, is in the 
diastolic position. The latter kind of action is the only one which is observed in frogs 
or mammals after subcutaneous administration of harmaline. 



272 THE PHARMACOLOGICAL ACTION OF HARMALINE. 

In the frog, after destruction of the central nervous system, harmaline constricts 
the blood-vessels when perfused through them. 

In mammals, small doses of harmaline cause a rise of blood pressure — always in the 
rabbit, and sometimes in the cat. The rise is due to contraction of the arterioles, and 
is accompanied by a diminution in the rate of the heart. Large doses produce a fall 
of blood pressure due to slowing and weakening of the heart. 

The slowing of the heart produced by an initial small dose is partly due to reflex 
stimulation of the vagus by the rise of blood pressure ; that produced by larger doses 
is independent of vagus stimulation, and due to an action on cardiac muscle. 

In frogs, paralysis of respiration is an early effect of lethal doses ; and in mammals, 
arrest of respiration, due to paralysis of the respiratory centre, is the chief cause of 
death from harmaline poisoning. In the latter there is frequently an initial stage of 
increased respiratory activity. 

Large doses of harmaline cause a fall of temperature in mammals ; the fall is 
slight with non -lethal doses, but may be profound in the case of a slowly lethal dose. 

Harmaline exerts a powerful action on uterine muscle in the direction of inducing 
sustained tetanic contraction. 

Harmaline can therefore no longer be regarded merely as a respiratory and 
convulsant poison. It differs from most alkaloids in that it does not exert, to the 
same extent as they do, a selective action on one kind of tissue. It attacks not only 
highly specialised tissues such as voluntary muscle, muscle of the heart, blood-vessels, 
and uterus, and cells of the central nervous system, but also less highly differentiated 
cells, such as pigment cells, protozoa (Raab), and ciliated epithelium (Jacobson). 

In this account of its pharmacology the actions of harmaline have been shown to 
resemble very closely those of another alkaloid, of which the above type non-selective 
action is also true, viz. quinine. As a pharmacological agent, harmaline ought to be 
grouped with quinine, and therefore with those substances which are conveniently, if 
somewhat indefinitely, termed " protoplasmic poisons." 

Considering the close resemblance in the pharmacological actions of harmaline and 
quinine, one is led to anticipate some corresponding similarity in their therapeutic 
effects. With this subject I hope to deal on a future occasion. 



( 273 ) 



XII. — The Pharmacological Action of Protocatechyl-Tropeine.* 
By Professor C. R. Marshall. 

(MS. received November 20, 1909 Read June 21, 1909. Issued separately January 7, 1910.) 

Protocatechyl-tropeine is one of a series of new tropeines investigated primarily 
with the object of determining the difference in pharmacological action between a 
lactone and the corresponding hydroxy -acid, f Its constitution is shown in the 
following formula : — 



HC 
HC 



COH 
C-OH 



CrL 



-CH- 



-CH 



CH 
CCOO-HC 



NCH Q 



CH„ CH- 



-CH 



Like most other tropeines, it paralyses the vagal nerve endings in the heart. It 
also diminishes the irritability of voluntary muscle and its myo-neural junctions, 
and paralyses the respiratory centre. 



Action on the Respiration. 

When O'Ol g. protocatechyl-tropeine hydrochloride is injected into the external 
jugular vein of an anaesthetised rabbit or cat the respiration is rapidly paralysed (fig. l). 
It generally ceases within ten seconds of the commencement of the injection, and may 
or may not recommence spontaneously. If it does not recommence, artificial respiration 
performed for ^-1 min. will fully re-establish it. After larger doses (e.g. 0'02 g.) 
it may be necessary to resort to short periods of artificial respiration three or even four 
times before the respiration is permanently established, and then it may remain 
somewhat shallower than at first. (In one animal in which the uncut sciatic or crural 
nerve was being stimulated at intervals, the respiration restarted spontaneously after 
the injection of this dose, but it failed again later.) Doses of 0*005 g. intravenously 
may also paralyse the respiration, but it recommences with a shorter period of artificial 
respiration than is the case after larger doses. Smaller doses produce only shallower 
and less frequent respirations, the effect being roughly proportional to the dose. The 
smallest effective dose is 0"001 g. intravenously. 

* I am indebted to Mr H. A. D. Jowett, D.Sc, for his kindness in providing me with this substance. Its 
preparation and characters have been described by him in conjunction with Mr Hann (Trans. Chem. Soc, vol. 89, 
p. 364). 

t Arch. f. exp. Path. u. Pharm., 1908, Supplem. Bd., p. 389. 

TRANS. ROY. SOC, EDIN., VOL. XLTII. PART II. (NO. 12). 41 



274 PROFESSOR C. R. MARSHALL ON 

Division of both vagi has relatively little effect on the respiratory action of 
protocatechyl-tropeine. In many cases the respiration seemed to recover less readily 
than in animals with uncut vagi ; and in a few cases the respiration did not recommence 
spontaneously, as it probably would have done if the vagi had been intact. I have 
observed the same effect after the administration of tetramethyl-ammonium chloride, 
a substance which acts on the respiration in a manner similar to protocatechyl-tropeine. 

The experiments were made on etherised animals. The blood-pressure was taken 
from the common carotid artery, and the injection made into the external jugular 
vein in cats and the facial vein in rabbits. The respiration was recorded in most 
experiments by means of a phrenograph with a tube connection to a tambour. 

In a normal animal the largest dose I have been able to give, namely, 0'045 g. 
intraperitoneally to a rabbit weighing 720 g., produced only dyspnoea. The duration 
of action of this dose and its influence on the frequency of the respiration is shown 
in the following table : — 



Time in minutes 
after injection 



10 20 25 30 40 50 60 90 120 



I 5 

No. of respirations I ^ M g7 153 lg5 135 m m 1Q5 7g g4 
per minute ) 

When injected into the dorsal lymph sac of winter frogs, 0*04 mg. per gramme 
body-weight caused slight slowing of respiration, but no other obvious symptoms ; 
0*08 mg. per gramme body- weight induced marked slowing of the respiration, which 
also took on a periodic character, groups of a bout six respirations being followed by an 
interval up to half-minute duration of no breathing ; there was also some muscular 
depression; 0"16 mg. per gramme body- weight caused almost complete cessation of 
respiration and muscular paralysis; and 0"24 mg. per gramme body-weight eventually 
produced complete paralysis of the respiration and death. As an example, the effect 
after the injection of 0*16 mg. per gramme body- weight may be taken. In this 
experiment the frequency of the respirations fell to one-half in ten minutes, to one- 
eighth in twenty minutes, and to one-twentieth in seventy-two minutes after the 
injection ; the frog was then pithed. 

Action on Vagal Endings. 

This was determined by observing on the blood-pressure the effect of electrical 
stimulation of a divided vagus before and after injection of the drug, or the effect of 
the drug after the injection of pilocarpine. When electrical stimulation was employed 
the procedure was as follows : the vagus was stimulated at intervals until a practically 
constant effect was obtained, the drug was then injected into the external jugular or 
facial vein, and the vagus again stimulated at intervals until an effect as near as possible 
identical with that previously obtained was reached ; or the injection was made during 
continued stimulation of the vagus (fig. 2). 



THE PHARMACOLOGICAL ACTION OF PROTOcATECHYL-TROPEINE. 



275 




2 c 



© g 

O CO 

.2 & 

ai .■» 



t, oeM 



3 -K 



~ r2 *= 



p4 ■- 

O C9 



" .2 



60 



& S 

S3 13 
1,H O 
CO £ 



CO £ O 

*3 S-^ CD 

2 * c? 

3 CO '" 
> -* 2 



S 13 



fl X 



&s 



0> 


T3 


- 




o 
o 


s 




Cd 


-M 


+3 

|3 


,d 


TJ 


o 


5" 


5 


'5b 


tc 
o 
G 
a 


CD 




















o 




CO 


33 




3 

bo 




3 




CO 


ce 


o 


s 


'S. 

CO 


n 

o 


tub 


CD 




o 


II 


13 


— 
o 


*. 


s 


CO 








G 


to 




o 






o 


CO 


o 


CD 


13 


1 


'a 


2 
o 




« 



w 



o 



a a 

+3 o 

co <N 

bCT3 






« .5 

s> pa 



o co 

-G <f 
CD " 






276 PROFESSOR C. R. MARSHALL ON 

By these methods 0"01 g. protoca techy 1-tropeine hydrochloride was found to 
paralyse the terminations of the vagus in the heart for five minutes. At the end of 
this period stimulation of the vagus generally produced a slight effect, and the effect 
gradually increased with succeeding stimulations until the former normal condition 
was reached, usually about twelve minutes after the injection. Occasionally the same 
strength of stimulus failed to induce any action, and consequently it had to be slightly 
increased. Thus, after the injection shown in fig. 2, stimulation with the secondary 
coil at 20 cm. produced no effect up to eight minutes after the injection, whereas 
stimulation with the coil at 16 cm. produced at this time a very good effect. I have 
obtained the same result with other tropeines. 

When pilocarpine is used as a stimulus the paralysis of the vagal terminations does 
not last so long. Thus the injection of O'Ol g. protocatechyl-tropeine hydrochloride, 
after a previous injection of O'Ol g. pilocarpine, totally abolished the pilocarpine effect 
for one and a half minutes only. The pilocarpine action then reappeared and in three 
minutes had reached its previous form. The shorter period of paralysis in these 
experiments as compared with that obtained when the vagus is stimulated electrically 
is due to the fact that pilocarpine and protocatechyl-tropeine are to some extent 
mutually antagonistic. 

Effect on Blood-Pressure. 

The commonest effect on the blood-pressure is a gradual fall, usually commencing 
after the effect on the respiration has begun, and continuing when the respiration 
spontaneously recovers, after this has become practically normal (cf. figs. 1 and 4). 
The fall of blood-pressure continues for 1^ — 3 minutes, rarely more, and then the 
pressure gradually returns to its previous height. It is not usually severe in extent 
in rabbits; after a dose of 0*01 g. intravenously it generally amounts to about one- 
third the normal blood-pressure. In cats the fall is often greater, and not infrequently 
is preceded by a slight rise. The extent of the fall in the last-named animals is shown 
in the following table, which expresses the result of two successive injections : — 

Minutes after injection ..... 3 9 

0004 g. Bp. in Mm. Hg 135 79 130 

0-002 g. Bp. in Mm. Hg 136 111 130 

Occasionally in cats a rise of blood-pressure has been the most prominent feature. 
This effect is seen in fig. 3, which illustrates the greatest rise of blood-pressure I have 
obtained with this substance. There was no fall below the normal subsequently. The 
vagal terminations in the heart were found to be paralysed between A and B, and 
to be again irritable between B and C. Two injections of O'Ol g. each were made in 
the same animal later, and in both instances a rise of blood-pressure to nearly the height 
of that shown occurred. The rise was more prolonged in each case, and the vagal 
endings were paralysed for a longer time than after the injection illustrated. The 
respiration was not graphically recorded in this experiment. 



THE PHARMACOLOGICAL ACTION OF PROTOCATECHYL-TROPEINE. 



277 



o 









w 


CO 




kO 


bC 






T3 


id 


03 


Cl 


en 






ago 

-a « 

o w 

o p 

gs 

£> c 

O =3 

C 09 

.5 

§ 05 
^ S 



I'-S 

3 -5- 



pq 



o c 



?3 +j 



o S 
I fee 

CO P*' 




w £ 



T3 


S3 


- 


03 


o 




o 
M 


to 

a 


o 






w 






M «H «i 



T3 t) 

a> it! 

CD 

> ii 



s ° 

o o 

CI ^ 



c » 

2 S 

d ca 

53 «= 

s » 

S3 ^3 

8 * 

; t- o> 

; =2 .£; 

I ^ w> 

' 45 § 

I 12 « 

§ ~ 

£ s 

r-T 'S 

I 2 

CB >a 

c -5s 



3 & 

o 

O <D 

,a a 



£ fa 

,_, o 

ccj o 



n w 
"53 a 
a< o 
p -_f3 ^ T3 
+" r2 J2 » 



&> S 



s a 

P.-J3 
bb£ 

3.1 

Si 

... in 

1 s 

C5 



T— ( 
S3 



&=( 



278 PROFESSOR C. R. MARSHALL ON 

When the respiration is paralysed and does not recover spontaneously the blood- 
pressure after a brief interval falls rapidly to a few millimetres. That this fall is 
mainly asphyxial is proved by artificial respiration, which quickly raises the pressure 
to the level it would probably have reached if respiration had recommenced. If the 
substance is injected after the administration of pilocarpine or during stimulation of the 
vagus a rise of blood-pressure always occurs. 

The fall in blood-pressure is probably cardiac in origin, since it is accompanied 
by a diminution in the frequency of the heart-beats ; and as the vagus nerve endings 
are paralysed, the action of the drug would seem to be on the muscular tissue of the 
heart itself. The rise of blood-pressure seen occasionally in experiments with cats 
is accompanied by an increase in the number of the heart-beats due to paralysis of 
the vagal terminations. The cardiac muscle in these cases is not depressed sufficiently 
to counteract the effect of paralysing the vagus, but that some depression is produced 
is shown by the rise of blood-pressure being less than that obtained with a correspond- 
ing physiological dose of atropine. Moreover, increase in dose leads to a somewhat 
smaller rise of pressure. Why this prolonged rise should occur in isolated cases is difficult 
to explain. Probably a certain concentration of the drug in the heart acting for a brief 
interval only is necessary, but the condition of the heart itself may also be a factor. 

Action on Voluntary Muscle and Motor Nerves. 

The activity both of motor nerves, or more probably their terminations, and of 
voluntary muscle, is depressed by this drug. In rabbits the nerve endings appear to 
be more susceptible than the muscle ; in frogs the muscle is at least as powerfully 
affected as the nerves. After an injection of 0'24 mg. protocatechyl-tropeine 
hydrochloride per gramme body-weight into the dorsal lymph sac of a frog, muscular 
weakness appeared within five minutes — the head began to droop and the animal 
became unable to turn over when placed on the back. A few minutes later the frog 
was quite flaccid. This condition and the paralysis of the respiration referred to 
previously was maintained during the period of observation (the frog died during the 
night). Although apparently completely paralysed, the frog occasionally made a slight 
voluntary movement, and a slight reflex was several times obtained on pinching a toe 
moderately strongly for some seconds. After 0*16 mg. per gramme body- weight was 
injected similar symptoms were observed. The head commenced to droop in twenty 
minutes and the animal gradually sank on the table. The reflexes, however, remained 
distinct, and when placed on the back an hour after the injection the frog recovered 
once, but not twice, from this position. The frog was pithed seventy-two minutes after 
the injection and the irritability of the sciatic nerves and gastrocnemii muscles tested. 
The sciatics reacted with the secondary coil at 35 cm., the muscles with the coil at 28 
cm. The injection of # 08 mg. per gramme body-weight produced only slight muscular 
depression. 



THE PHARMACOLOGICAL ACTION OF PROTOCATECHYL-TROPEINE. 279 

On immersing a nerve-muscle preparation (gastrocnemius) in a 1-1000 solution of 
protocatechyl-tropeine hydrochloride in 0*6 per cent, sodium chloride and stimulating 
(one Daniell's cell) the nerve and muscle alternately every three minutes, the 
contractions were found to fall with the same regularity in each case, and both muscle 
and nerve became inexcitable forty-five minutes after immersion. 

In a rabbit no muscular paresis was observed after 062 g. p. kg. intraperitoneally. 
Depression but not always paralysis of the nerve endings is obtained after intravenous 
administration to an anaesthetised animal. This is shown in fig. 4. The rabbit was 
prepared in the manner described, the right crural nerve was laid bare, and a hook was 
put into the right foot and connected by means of a thread working over pulleys to a 
weighted lever. After a practically constant effect had been obtained from stimulation 
of the nerve (one accumulator cell; secondary coil 12 cm.), 001 g. protocatechyl- 
tropeine hydrochloride was injected into the right facial vein. The contraction of the 
muscles produced by stimulation of the nerve gradually diminished until it disappeared 
two minutes after the commencement of the injection. Two and a half minutes later 
the limb muscles again responded to stimulation of the nerve, but the contractions did 
not assume their previous (normal) form until ten minutes after the injection of the 
drug. The tracing reproduced records a second injection of the drug ; the same dose 
was administered at 11.19 and produced a similar effect. 

In a second experiment, in which both the right crural and sciatic nerves were 
exposed, it was found that the muscles, when stimulated directly, contracted in the 
neighbourhood of the electrodes at a time when the nerves were inexcitable. It would 
seem, therefore, that in the rabbit the nerve endings are more susceptible to this 
compound than the muscular substance. 

An interesting point is the want of correlation between the respiratory and the neuro- 
muscular effects ; at a time when the latter are most manifest the respiration is 
practically normal. The explanation is probably to be found in the stimulus of the 
respiratory centre being physiologically more powerful than the electrical stimulus 
used, and thus able to overcome any tendency towards paresis of the respiratory 
muscles. As bearing upon this point, evidence was obtained showing that the paralysis 
of the muscles and nerve endings was only relative, i.e. only for minimal or moderate 
stimuli. The effects on the blood-pressure and the neuro-muscular system, on the 
other hand, are correlative, the fall of blood-pressure being mainly, if not solely, due 
to an action on the cardiac muscle. 

Remarks on the Action of the Drug on the Respiration. 

Rapid cessation of respiration in the expiratory phase has been described by 
Tappeiner* after the intravenous injection of methyl-phenyl-isoxazol-methochloride, 
diphenyl - methyl - pyrazol - methochloride, dimethyl - phenyl - pyrazol - methochloride, 

* Arch. f. exp. Path. u. Pharm., xxxvii., p. 325 [1896]. 



280 



PROFESSOR C. R. MARSHALL ON 



and tetra-methyl-ammonium chloride ; and by Pohl* after the intravenous injection of 
papaverin-methochloride, papaveraldin-methochloride, and papaverinol-methochloride. 
None of these substances produced the effect when administered hypodermically. 

As the result of his experiments, Tappeiner came to the conclusion that methyl- 
phenyl-isoxazol-methochloride stimulated the terminations of the trigeminal nerve, and 
that the effect produced was, in reality, a Kratschmer-Hering reflex. Thus he found 
that cocainisation of the nasal mucous membrane prevented the respiratory paralysis ; 
and that while cocaine administered intravenously also antagonised the action of the 
isoxazol, the dose required was larger than when the cocaine was applied directly to the 
nose. The action of tetramethyl-ammonium chloride and the other substances investi- 
gated by him was explained in the same way. Iodlbauer,'!' who investigated tetramethyl- 
ammonium chloride more minutely under Tappeiner's direction, came to the same 
conclusion. Pohl, however, still obtained the characteristic stoppage of the respiration 




Fig. 5. — Effect of 0"01 g. protocatechyl-tropeine hydrochloride intravenously on decerebrate rabbit with both fifth cranial 
nerves cut. Tracing of respiration taken by connecting one limb of tracheal cannula to a tambour ; other limb left open. 
Time in seconds. 

after cocainisation of the nose and after division of the ophthalmic branch of the fifth 
nerves, both in the case of papaveraldin-methochloride and of tetramethyl-ammonium 
chloride, and he therefore maintains that the action is a central one. The action of 
protocatechyl-tropeine is also upon the centre. The typical effect on the respiration is 
still obtained when the brain above the pons is excised and both the fifth cranial nerves 
are cut within the cranium (fig. 5). Moreover, the same action is obtained in cats 
(fig. 1), animals in which the typical Kratschmer-Hering reflex is not present. And 
since the effect on the neuro-muscular system is not coincident in time with the effect 
on the respiration, the cause of the latter would seem not to be peripheral. It has, 
however, been shown that the motor nerve endings in the diaphragm may be affected 
differently, either as regards time or degree, from those of many other voluntary 
muscles, and it therefore seemed desirable to determine whether the diaphragm played 
any part in producing the respiratory paralysis. Accordingly, in an anaesthetised 



* Arch. Internat. de Pharmacol., xiii., p. 479 [1904]. 



t Ibid., vii., p. 183 [1900]. 



THE PHARMACOLOGICAL ACTION OF PROTOCATECHYL-TROPEINE. 



281 



animal, both phrenic nerves were cut and the respiration registered by a connection 
with the tracheotomy tube. On injecting 0"01 g. protocatechyl-tropeine hydrochloride 
into the external jugular vein the respiration ceased thirteen seconds after the 
commencement of the injection (fig. 6). 

In connection with the production of temporary cessation of the respiration, the 
question of dose is of paramount importance. The concentration of the drug producing 
this effect in the blood supplying the medulla can apparently vary only within narrow 
limits. Hence not only the amount of drug injected but also the time occupied by the 
injection is of moment. This explains why the effect does not result from hypodermic 
or intraperitoneal administration. It is probably owing to this fact that the effect has 
not been observed more frequently. It can certainly be produced by a number of 




innririiiriiwinji^^ 



o-*c<. z% 



Fig. 6. — Effect of O'Ol g. protocatechyl-tropeine hydrochloride on respiration of rabbit with both phrenic nerves divided. 
Tracing taken as fig. 5, but with air-exits closed. Time in seconds. 

drugs, but the doses of those which induce it fall within very narrow limits. In many 
of these cases the effect is probably merely that of a protoplasmic poison, and the 
respiration is most obviously affected because the respiratory centre is the most 
sensitive of the medullary centres to this class of substances. This hypothesis, however, 
does not seem sufficient to explain the actions of the substances already mentioned. 

In a preliminary report on the physiological action of pukateine, an alkaloid 
isolated by Aston * from pukatea bark, Malcolm says : " Intravenous injection of 10 mg. 
into an anaesthetised rabbit caused sudden stoppage of respiration without convulsions : 
the heart continued to beat vigorously for a long time afterwards." As Mr Aston 
had given me some years ago a specimen of the mixed alkaloids isolated by him from 
pukatea bark, I was able to investigate this point. The alkaloids were first converted 
into the hydrochlorides. On injecting 0*01 g. intravenously (time occupied by 
injection, seven seconds) into a rabbit (wt. 2100 g.) the respiration was not 

* Annual Rep., New Zealand Dept. of Agriculture, 1908, Chemistry Division, p. 226. 
TRANS. ROY. SOC. EDIN., VOL. XLYII. PART II. (NO. 12). 42 



282 PROFESSOR C. R. MARSHALL ON 

paralysed ; it began to get shallower ten seconds after the commencement of the 
injection, and reached its lowest point five seconds later, from which time it very 
gradually returned to the normal. The blood-pressure commenced to fall before the 
respiration was affected, and fell from 87 mm. before the injection to 58 mm. twelve 
seconds after the commencement of the injection. From this point it gradually rose in 
three minutes to 107 mm. The effect of a subsequent injection of 0"02 g. caused 
almost complete cessation of respiration for half a minute, and then a very gradual 
return to normal breathing. The fall in the respiratory curve, however, was less 
abrupt, the blood-pressure was more quickly and more powerfully influenced, and the 
return of normal respiration was later than occurs after protocatechyl-tropeine. 

The other synthetic tropeines I have investigated do not distinctly influence the 
respiration in doses which depress the vagal endings, but doses of both atropine and 
homatropine may be found which will cause temporary cessation of the respiration. 
The dose of atropine necessary for rabbits is about O'OOOl g. per kg. body-weight 
intravenously. After this dose, however, the return of normal respiration is less 
rapid than after the administration of protocatechyl-tropeine. Homatropine, in doses 
of 0'005 g. per kg. body-weight intravenously, produces a more transient effect on the 
respiration and a greater effect on the blood -pressure than atropine in the dose 
mentioned. I have also succeeded in producing temporary stoppage of the respiration 
with the allied alkaloid cocaine, and with /3-eucaine and other local anaesthetics. I 
failed, however, to produce the effect with ammonium hydroxide. A certain degree of 
recovery may, with appropriate doses, be obtained after intravenous administration of 
this substance, but this recovery is not maintained, and the respiration fails again unless 
artificial respiration is performed. The best effect I have obtained was from the 
injection of 0*025 g. ammonium hydroxide per kilogramme body-weight. Larger 
amounts of ammonium hydroxide may be followed by a few spontaneous respirations, 
but they paralyse the respiration permanently unless it is restored by artificial means. 
Doses less than O'Ol g. per kilogramme generally increase the respiratory movements; 
above this dose the movements are diminished. 

Comparison op Effects on Blood-Pressure. 

All the drugs previously mentioned, in doses which temporarily abolish the respira- 
tion, diminish in rabbits the frequency of the heart-beats. The effect is greatest in 
the case of methyl-phenyl-isoxazol-metbochloride and tetramethyl-ammonium chloride, 
and least in that of the tropeines. In most cases there is also a fall in blood-pressure. 
Methyl-phenyl-isoxazol-methochloride, however, notwithstanding the marked slowing of 
the heart produced by it, causes, after a preliminary slight fall, a rise of blood-pressure, 
which is due, according to Tappeiner,* to stimulation of the medullary centres. The 
slowing of the heart and fall of blood-pressure produced by tetramethyl-ammonium 

* hoc. cit. 



THE PHARMACOLOGICAL ACTION OF PROTOCATECHYL-TROPEINE. 283 

chloride, and probably also by the alkaloid pukateine, are due mainly to stimulation of 
the vagal terminations. According to Iodlbauer,* the effect of tetramethyl-ammonium 
chloride is less after section of both vagi, and consequently he concludes that its action 
on the circulation is partly, if not mainly, central. He also obtained a marked rise of 
blood-pressure when this drug was injected after atropine, a result he attributes to 
stimulation of the vasomotor centre. I have obtained a rise of blood-pressure on cutting 
both vagi a few minutes after the injection of tetramethyl-ammonium chloride, but have 
not observed any difference on the frequency of the heart-beats after injection of this 
substance whether the vagi were divided or not. After atropine had been given, the 
usual slowing of the heart did not occur, nor was there any rise of blood-pressure. 

The slowing of the heart's frequency and the fall of blood-pressure produced by the 
papaverin derivatives investigated by Pohl are attributed by him to the stoppage of 
the respiration. He does not, however, refer to any investigations bearing on this point 
in his paper, and he gives no tracings from which the action on the circulation might 
be inferred. 

The action of protocatechyl-tropeine on the circulation differs from that of most 
substances producing temporary paralysis of the respiration, in that it paralyses the 
vagal endings within the heart. Hence the diminution in the frequency of the heart- 
beats and the fall of blood-pressure are less than are produced by most members of this 
class of respiratory paralysants. As previously stated, the fall of blood-pressure caused 
by these tropeines is probably due to a depressant action on the cardiac muscle, an 
action which in all probability is present in the case of the other substances also. 

From these statements it is evident that no uniformity, such as exists in the case of 
the respiration, characterises the action of these different substances on the circulation, 
and this further supports the statement made previously that their action on the two 
systems is not correlated. 

Remarks on the Relation between Chemical Constitution and 
Pharmacological Action. 

Protocatechyl-tropeine, as previously stated, was one of a series of tropeines in- 
vestigated for the purpose of determining the difference between a lactone and the 
corresponding hydroxy -acid. As it is not a lactone it was used as one of the control 
substances, and it was found that, unlike pilocarpine and the lactone-tropeines, it does 
not lose its action when a molecular quantity of caustic alkali is added to its aqueous 
solutions. Even the addition of distinct excess of caustic alkali and prolonged standing 
do not influence its pharmacological action to any appreciable degree. It is evident, 
therefore, that the replacement of the hydrogen of the phenolic hydroxyls by an alkali 
metal is of no pharmacological importance. 

A point of interest on which the action of protocatechyl-tropeine throws a little 

* Loc. cit. 



284 PROFESSOR C. R. MARSHALL ON 

light is the part which the constituent groupings play in the pharmacological action of 
tropeines. Buchheim,* LADENBURG,t and others have found that various aromatic 
acidic groups could be attached to the base tropine, and an atropine-like action 
be obtained ; whereas Crum Brown and Fraser,| and more recent observers, have 
shown that alterations in the tropyl radical generally produce profound changes in the 
pharmacological action of a tropeine. This suggests that the fundamental action of 
these tropeines depends on the tropyl radical, and that the interaction with the so- 
called nerve endings occurs through the mediation of the acidic grouping. Whether the 
acidic group has any other action than a connecting or haptophoric one with the tissues 
is difficult to decide. Many of the tropeines, in so far as they have been investigated, 
differ little except quantitatively in pharmacological action from atropine, and it is open 
to assume that in these compounds the acidic groupings are purely haptophoric in 
character, since their different quantitative effects might be explained by differences in 
the power with which the acid group is able to attach the tropyl group to the nerve 
endings. Gottlieb, § however, showed that certain tropeines act very differently from 
atropine — acetyl-tropeine causes convulsions and succinyl-tropeine paralysis, and neither 
affects the vagus endings or the pupil ; and in view of the comparatively simple nature 
of the acid groups of these compounds, it seems questionable whether any distinctive 
action can be attributed to the constituent groupings of tropeines. In the case of 
protocatechyl-tropeine we have a compound which differs materially in action from 
other tropeines, yet most of its actions can be demonstrated with atropine. But while 
it acts much less powerfully than atropine upon the vagal endings in the heart, it acts 
more powerfully on voluntary muscle, and has an action on the respiratory centre which 
atropine shares to a relatively slight degree. It may be that these differences might be 
explained by the presence of different receptive substances in the different tissues, but 
this view would scarcely explain why, in small doses, atropine exerts a stimulating effect 
on the respiratory centre, and protocatechyl-tropeine no such stimulant action. Un- 
fortunately, it is not possible to determine the action of the two constituent groups of 
these compounds by investigating the products of their hydrolysis. It was early 
shown by FraserU that tropine and tropic acid produce no atropine-like action, and it 
was consequently not surprising to find that simple protocatechyl compounds possess no 
action like that of protocatechyl-tropeine. The substances I investigated were proto- 
catechyl-aldehyde,|| methyl -protocatechyl-aldehyde (vanillin), and di-methyl-proto- 
catechyl-aldehyde (piperonal). Intravenous injections of these compounds produced no 
effect on the respiration resembling that produced by protocatechyl-tropeine. There is 
thus no evidence leading us to assume that the actions of protocatechyl-tropeine can be 
attributed with any certainty to the two constituent groupings individually. 

* Arch.f. ezp. Path. u. Pharm., v., p. 463 [1876]. t AnnaUn d. Chemie, Bd. 217, p. 82 [1883]. 

X Trans. Roy. Soc. Edin., xxv., p. 693 [1869]. § Arch. f. exp. Path. u. Pharm., xxxvii., p. 218 [1896]. 

% Proc. Roy. Soc. Edin., 1869, p. 558. 
|| Prepared by acting on piperonal with phosphorus pentachloride. 



THE PHARMACOLOGICAL ACTION OF PROTOCATECHYL-TROPEINE. 285 

All the substances previously investigated which have been shown to paralyse 
temporarily the respiration are quaternary bases. The tertiary bases corresponding to 
these did not, where tried, affect the respiration. Protocatechyl-tropeine, however, is a 
tertiary base, and chemically has nothing in common with the other substances pro- 
ducing the same respiratory effect. But it possesses another physiological action 
common to many of them, namely, the power of paralysing the motor nerve endings to 
voluntary muscle. It is questionable, however, if this connection is more than a co- 
incidence, since trimethyl-ammonium chloride, triethyl-ammonium chloride, and 
tetraethyl-ammonium chloride, which paralyse motor nerve endings, have no paralysing 
action on the respiratory centre ; * and the papaverin bases which do produce re- 
spiratory paralysis do not paralyse motor nerve endings. 

Summary. 

Protocatechyl-tropeine paralyses the vagal endings in the heart, but is much less 
powerful than atropine, or even homatropine. 

It depresses muscular activity, especially in frogs, and interferes with the con- 
ductivity of motor nerves, or more probably the myo-neural j unctions, in rabbits. 

In certain doses, it paralyses temporarily the respiration, and causes a gradual fall 
of blood-pressure ; but the two effects are not apparently correlated. In cats, a rise of 
blood-pressure sometimes occurs, owing to the paralysis of the vagal endings. 

The temporary stoppage of the respiration is due to paralysis of the respiratory 
centre, and not to a peripheral action. 

No evidence indicating separate pharmacological actions of the two component 
groupings of the compound was obtained. 

* Tappeiner, loc. cit. 



TRANS. ROY. SOC. EDIN, VOL. XLVII. PART II. (NO. 12). 43 



( 287 ) 



XIII. — The Pharmacological Action of Tutu, the Toot Plant of New Zealand. 

By Professor 0. R. Marshall. 

(MS. received November 20, 1909. Read June 21, 1909. Issued separately January 29, 1910.) 



CONTENTS. 



PAGE 

Introduction 287 

Active Principle 288 

Previous Pharmacological Investigations . . 289 

General Effect of Tutin 290 

on Rabbits 290 

on Guinea-pigs 294 

on Frogs 294 

on Trout and Tadpoles .... 295 

Action on Medullary Centres .... 295 



The Convulsions 

Influence of Anaesthetics on the Convulsions 
Muscles affected by the Convulsions 
Seat of Origin of the Convulsions 
Action on the Higher Centres 
Influence on Temperature . 
Effect of Decomposition Products 
Comparison with other Convulsants 
Summary 



PAGE 

300 
302 
304 
307 
309 
310 
312 
312 
314 



Introduction. 

Under the name Tutu, three species of Coriaria — C. ruscifolia (L.), C. thymifolia 
(Humb. & Bonp.), and C. angustissima (Hook.)* — are usually included. Other species 
have been described, but they appear to be varieties only. All the forms are poisonous, 
and in times past this plant has proved a serious hindrance to the rearing of stock in 
New Zealand. LiNDSAY.t in the course of a tour through the country during the latter 
part of 1861 and the earlier months of 1862, "was everywhere struck by the abundant 
evidences of the devastation produced among flocks and herds from their feeding on the 
toot plant" ; and he further states that " he seemed a fortunate farmer or runholder who 
had not lost more than 25 per cent, of his stock from toot-poisoning ; while in some 
instances the losses were so high as 75 per cent." 

In order to gauge the extent of its ravages at the present time, a circular letter, dated 
25th July 1900, was sent by Professor Easterfield and Mr Aston to every medical 
man in New Zealand, and to such laymen as they thought might take an interest in the 
matter. Very few replies were received (4 from medical men, 10 from lay men), | and 
none of them indicate the extent of the malady. On the whole, the correspondents 
corroborate the observations of Lindsay (I.e.). The following account summarises the 
statements made in the letters and in the extracts from the lay press submitted to me. 

* G. ruscifolia is commonly known as " tree-toot." It grows to a variable height, sometimes to 6-8 metres, and 
has glossy, succulent leaves and shoots. G. thymifolia or " ground-toot : ' usually grows to 30-50 cm. in height, and has 
small leaves and flowers. G. angustissima is herbaceous, and grows about 50 cm. high. Its slender branches terminate 
in very fine branchlets which carry numerous leaves, thus giving the plant a fern-like appearance. It is compara- 
tively rare. 

t "On the Toot Plant and Poison of New Zealand," Brit, and For. Med.-Ghir. Rev., July 1865, p. 153. 

J These were kindly sent to me by Mr Aston. 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART II. (NO. 13). 44 



288 PROFESSOR C. E. MARSHALL ON 

The plant seems to be most dangerous in spring. At this season it puts forth thin, 
succulent shoots before other vegetation appears, and it is consequently liable to be 
eaten in large quantities as soon as the animals are turned out to graze. Later in the 
year, when fodder is abundant, poisoning is less common and the symptoms less severe. 

The symptoms observed after tutu is eaten are very similar in both cattle and 
sheep. Unsteadiness, stupor, and convulsions are the chief, and they usually occur in 
this order. Any one of these may be predominant, but the convulsions are the most 
characteristic. After being " tooted " the animals may show no symptoms until they 
are aroused, when unsteadiness and staggering expose the condition. The animals, it is 
said, frequently run about as if wild, often against obstacles, and if any water is near at 
hand they may rush into it and be drowned. If such an accident does not occur, 
convulsions of a clonic type develop, and the animals die in these or during the period 
of exhaustion which follows. 

Other animals are affected in a similar manner. Fowls have been poisoned by 
eating the seeds, and an elephant * by eating the young shoots. Pigs and rabbits t are 
said to be immune, but the statement is untrue. This was proved for the pig by 
Mr J. A. Gilrtjth (v. p. 290), and for the rabbit by myself. Men, and especially 
children, have frequently been poisoned by eating the fruits (so-called " berries"). The 
juice of these is innocuous, and was made into a beverage by the Maoris ; the seeds, 
however, are poisonous. The main symptoms noted have been vomiting, giddiness, 
stupor or excitement with delirium, convulsions, and coma ; but cases appear to have 
varied considerably in the clinical picture they have presented. A notable effect is 
loss of memory : in most of the cases of recovery described, there has been no remem- 
brance of eating tutu, or of the convulsions and other symptoms which followed. 

Active Principle. 

The earliest recorded chemical investigation of the toot plant was made by Mr 
W. Skey,J Analyst to the Geological Survey of New Zealand, and was published in 
1869.§ By exhausting an alcoholic extract of the seeds of C. ruscifolia with ether, he 
obtained a pale green viscid oil, five minims of which, administered to a cat, caused 
almost immediate vomiting, followed in half an hour by uneasiness and convulsive 
twitches of the ears and eyes accompanied by forward jerking of the head. These 
symptoms culminated in a convulsive fit about one hour after the substance was 
administered. The animal eventually recovered. He concluded that the active 
principle existed in the oil, if it was not the oil itself ; and he conceived the idea that 

* Trans. N.Z. Instil, vol. ii. p. 399 [1869]. 

+ Ibid., vol. iii. p. 242 [1870]. J Ibid., vol. ii. p. 153 [1869]. 

§ Dr Laudkr Lindsay (I.e.) mentions that Dr Murray Thomson of Edinburgh undertook a chemical investigation 
of the specimens brought home by him, but owing to Dr Thomson's removal to India the research was not completed. 
Skky also states that "a great many experiments have from time to time been made upon the toot plant with the 
object of extracting the formidable poison known by sad experience to exist therein"; but as no other mention is made 
of these, they had probably ended in failure. 



THE PHARMACOLOGICAL ACTION OF TUTU. 289 

the oil was a fixed oil, and consisted of esters of peculiar fatty acids, to one or more of 
which acids he ascribed the poisonous effects obtained. In the following year Hughes * 
endeavoured to show that the active principle was an. alkaloid ; but the crystalline 
principle he isolated was not properly examined, and his experiments are otherwise 
unconvincing. Nothing further was done until 1900, when Easterfield and Aston,! 
by extracting a purified decoction of the plant with ether, obtained a crystalline 
glucoside from each of the three kinds of tutu. This substance, which they termed 
tutin, and to which they ascribed the formula C 17 H 20 O 7 , has proved to be the active 
principle of the plant. It was sent to me for pharmacological investigation, and I 
desire at this point to express my indebtedness to them for the material with which 
they have supplied me. 

Other substances were isolated from the three species of Coriaria investigated : 
quercetin from C. ihymifolia ; acetic, gallic, succinic, and a few other well-known acids 
from C ruscifolia ; and a volatile acid, C 8 H 8 4 , which has not been identified, from 
C. angustissima ; but these are obviously not the cause of the physiological activity of 
the toot plant. 

Previous Pharmacological Investigations. 

Apart from the published descriptions of accidental cases of poisoning in man and 
animals, and the solitary experiment of Skey already mentioned, our knowledge of the 
action of tutu is limited to a few experiments made by Mr AchesonJ and to an 
investigation of Dr W. Ledingham Christie. § 

Acheson administered an aqueous extract prepared by Hughes to dogs and cats, 
and noticed, after giving 15 grains or more, increased frequency of respiration, with 
slight twitchings of the extremities followed by severe attacks of convulsions. He 
found that the extract rapidly decomposed, and after a few days gave practically no 
pharmacological action, and that lime decomposed the poison and seemed to act as an 
antidote. 

Christie's investigation included some observations on the effect produced by 
different doses of a crude extract, and some experiments to determine the mode of 
action of the drug. After the administration of small doses of an aqueous extract by 
the mouth or hypodermically to dogs and cats, he observed salivation and apparent 
nausea, and after somewhat larger doses vomiting also. Still larger amounts (corre- 
sponding to 40-50 grains of the leaf) caused in addition, and commencing within half 
an hour, " twitching of the smallest muscles, as those of the eyes, nose, and ears. Then 
the muscles in front of the neck began to jerk, the fore paws began to follow suit, 
and soon the hind limbs and tail were affected." Early and late dyspnoea, erection of 
hairs, and increased frequency of micturition and defecation were noted. The vomiting 
he believed to be partly gastric but mainly medullary in origin ; the early dyspnoea 

* Trans. N.Z. Instit., vol. iii. p. 237 [1870]. t Trans. Ghern. Soc, vol. Ixxix. p. 120 [1900]. 

{ Trans. N.Z. Instit, vol. iii. p. 241 [1870]. § N.Z. Med. Journ., 1890, July and October. 



290 



PROFESSOR C. R. MARSHALL ON 



he attributed to the nausea, and the late dyspnoea to the convulsions. Rabbits he 
found insusceptible ; but the method of administration adopted — injection of an ethereal 
solution, at first hypodermically, then into the pleural cavity, and later into the 
peritoneal cavity — in the two experiments mentioned was unfortunate. And his 
experiments on birds are similarly vitiated. 

He endeavoured to localise the action of tutu by the following experiments. After 
obtaining convulsions from an intrapleural injection of two minims of oil of tutu, he 
sliced off the grey matter of the left motor area, and noted that when the effect of the 
anaesthetic had passed away the movements were limited to the right fore and hind 
paws. From this experiment he concluded that " when the cortex is absent there are 
no convulsions in the corresponding muscular area/' In a second experiment he 
injected an ethereal solution of oil of tutu subdurally in the right parietal region, and 
obtained symptoms which were undoubtedly due to the ether injected and the 
increased intracranial pressure produced ; but he obtained also, as late symptoms, 
twitchings of the left eyelid and later of the left fore and left hind limbs. Eventually, 
the movements became more symmetrical. In a third experiment the ethereal solution 
was applied to a (decorticated ?) portion of the left cortex, and was also " inserted deep 
towards the base posteriorly." " Blindness and impaired intellectual perceptions " were 
marked, and a convulsion subsequently supervened. Convulsions are said not to 
occur below a section of the spinal cord, but no details of the experiment are given. 
From these experiments he concludes that " the cortex of the motor area is the chief 
part affected ; but the sensory and cerebellar as well as the optic lobes, etc. , or corpora 
quadrigemina, seem to be irritated, for movements of the eyes occurred in some cases." 
Earlier in the paper he doubts if the basal ganglia take any part in producing the 
convulsions. 

General Effect of Tutin. 



The effect of tutin was tried, soon after its isolation, on a pig by Mr A. J. Gilruth, 
M.R.C.V.S., and on two cats by Mr A. R. Young, M.R.C.V.S. The pig (35 lbs.) 
received two grains dissolved in water with its food. In half an hour spasmodic closure 
of the mouth was noted, and somewhat later accelerated respiration and vomiting, 
which increased in severity until convulsions set in. The animal died four and a half 
hours after the administration. The cats which were given tutin in solution suffered 
mainly from convulsions followed by coma. Death occurred in forty minutes. 

My own observations have been made on rabbits, guinea-pigs, frogs, tadpoles, and 
young trout. In all, the effects produced were very similar, so that, taking into 
consideration the accidental poisoning observed in many of the higher animals, it is 
evident that the toot plant is a poison to all classes of vertebrates. 

On Rabbits. — In rabbits, after a lethal dose there is a variable period of quietude, 
accompanied by a diminution in the number of the heart-beats and an increase in the 
frequency of the respirations, which is followed by tonic and clonic convulsions. The 



THE PHARMACOLOGICAL ACTION OF TUTU. 291 

earlier convulsions usually commence in a tonic spasm. The animal presents a 
frightened appearance, the muscles of the limbs increase in tone and become rigid, 
and opisthotonos develops. This tonic spasm terminates in a clonic one, which in 
turn ends in exhaustion. During the convulsion, respiration is absent, and it may not 
return until the period of exhaustion has lasted for two minutes or more. A few 
minutes later the animal sits up and appears normal. This alternation of convulsions 
and exhaustion with a normal appearance may occur many times ; but after a few 
convulsive attacks, clonic movements of the limbs, usually accompanied by retraction 
of the head, generally become constant. Salivation, dilatation of the pupils and of 
the vessels of the ears, are frequent symptoms. Death may occur in the first or a 
subsequent convulsion, or may result from exhaustion. 

After non-lethal doses the convulsions are generally less severe but more character- 
istic. There is, at first, usually an appearance of clinging to the ground, often 
accompanied by muscular tremors, and the gait becomes hesitating and spasmodic ; 
but the convulsions themselves commence with twitches about the eyes or mouth, or 
with erection or retraction of the ears, and extend to the whole head and fore part 
of the body. Frequently these movements rapidly terminate, especially in the earlier 
attacks, in a tonic convulsion ; but later the convulsions are almost solely clonic. At 
this period, most frequently in animals which have been given a relatively small dose, 
a curious series of movements often occurs. The animal commences to push backwards 
with its forefeet, the hind parts being kept fixed, until it may, when the whole move- 
ment is carried out, perform a somersault over its tail. The movement may be 
abandoned at any point, and the animal may remain in the constrained position or 
assume a normal attitude. The movement may also be preceded, or accompanied in 
its early stages, by clonic spasm of the face or head, and it may terminate after the 
somersault in a convulsion. Christie mentions somersaulting round a transverse axis 
in the protocol of one of his experiments on cats. 

Tooted rabbits frequently lay, often for long periods, with the hind limbs more or 
less outstretched and with toes outspread as in the final stages of a spring. This 
position was so common that I termed it " characteristic position." 

Doses which do not produce convulsions have an effect on the medullary centres 
which will be described later. 

The lethal dose* for rabbits is about 0'0025 grm. tutin per kg. body-weight 
administered by hypodermic injection. After such a dose, death usually occurs from 
exhaustion. The lethal dose cannot be determined with exactitude, as it varies some- 
what with external influences. Thus a dose which is usually non-lethal may be 

* In a "Report on the Physiological Action of Tutin" by Dr Fitchett and Professor Malcolm, published in the 
Annual Report of the New Zealand Department of Agriculture for 1908, which has reached me since the above was 
written, the minimal lethal dose of tutin by hypodermic injection (unless otherwise stated) is given in mg. 
per kg. body-weight as follows : — cats, 0'75 ; guinea-pigs, 2 ; rabbits, 2 - 5 ; lizards, 3-4 ; birds, 10'5 (by oral adminis- 
tration) ; frogs, 10 - 5 ; fishes, 50 (when immersed in the solution). An extended report by Fitchett and Malcolm has 
appealed as this paper is passing through the press (see note at end). 



292 



PROFESSOR C. R. MARSHALL ON 



converted into a lethal one by turning the animal on its back a few times, thereby 
inducing convulsions. 

The minimal dose causing convulsions in rabbits is about O'OOl grm. per kg. body- 
weight. Larger doses may be given with impunity if the animal remains quiet, and 

Experiment I.— 0002 grm. Tutin per kg. Rabbit. Weight, 585 grm. 



Time. 



12.19 
12.29 
12.38 

12.44 

12.49 
12.59 

1.4 



1.27 



1.33 



1.39 

1.46 
1.57 

2.1 

2.8 

2.14£ 

2.25J 

2.55 

3.0 
3.2 
3.10 
3.30 
4.0 
Next 
morning. 



Observations. 



- 58 c.c 1 in 500 tutin solution injected under skin of back. 

Lay down and stretched hind limbs out. Respiration very rapid. 

Moved backwards somewhat spasmodically ; on resting, fore limbs commenced to slide for- 
wards. Respiration rapid. 

Occasional slight noddings of head. 

Head-nodding more marked. Tremors affecting whole foredialf of body. 

Convulsion, affecting almost solely right side of face and head — retraction of head with 
rotation to right, closure of right eye. Quickly recovered. 

Mild tremors over whole body. Moved about very carefully for a short time, and after- 
wards assumed a crouching attitude as if about to spring. A few spasmodic movements 
forwards were made later, but the same crouching position was maintained for fifteen 
minutes. Occasionally tremors over the whole body occurred. Pupils dilated. 

Clonic convulsion, mainly left-sided, and limited to fore part of body — the animal stood up 
on its fore limbs, the left ear was drawn back, the right remaining erect ; both sides of 
head, except the eyes, were next affected with coarse clonic movements. Convulsion 
lasted one and a half minutes. Slight salivation present. 

Another mild clonic convulsion ; limited at first to the mouth, then affecting the whole head 
(30" after commencement), and finally involving the hind feet (90" after commencement), 
which, however, only moved slightly. Rotation of head to left occurred (120" after 
commencement). Respiration, which had been markedly quickened, became slower. 

Laid on venter ; fore feet pushed out in front ; ears retracted. Almost constant tremor. 

A third severer convulsion lasting 135". At one time slightly right-sided. Salivation 
continues. 

An attempt to lick a part on the right side produced clonic movements which ended in a 
convulsion. Marked winking occurred in this. 

Clonic convulsion. In this convulsion the animal fell over on its side, and then showed very 
marked clonic movements — the legs moving very rapidly, as in running. 

Similar clonic convulsion. 

Convulsion affecting mainly the fore part of the body, especially the mouth and head. The 
animal repeatedly fell over on its side, but recovered its position each time. The con- 
vulsion lasted three minutes. 

Convulsion similar to last. 

Quiet since last note. Suddetdy made several high springs, and entered into a tonic convul- 
sion (slight opisthotonos) : a transient rest was followed by clonic movements (running, 
biting, etc., movements). Head retracted in the main. Salivation. 

Clonic movements of whole body have continued without cessation since last note. There 
has been an occasional spasmodic cry. 

Very transient tonic contractions. 

Tonic contractions of longer duration ; several cries. 

More powerful clonic movements; more frequent cries, with rapid respirations. 

Improving ; sitting up better. 

Convulsion ; fell on side, and exhibited running movements. Afterwards gradually improved. 

Practically well. 



probably smaller doses will induce convulsions if the animal is repeatedly stimulated,* 
but below this point the medullary effect of tutin is generally alone apparent. 

* It is well known in New Zealand that tooted sheep may not show any abnormal symptoms if left alone, but if 
driven they become excited and enter into convulsions. 



THE PHARMACOLOGICAL ACTION OF TUTU. 



293 



The course of the intoxication in rabbits after different doses of tutin is shown in 
the following protocols. Experiment I. gives the effect of a nearly lethal dose, and 

Experiment II. — O'OOl grm. Tutin per kg. Rabbit. Weight, 950 grm. 



Time. 


Heart- 


Respira- 


Tempera- 


Observations. 




beats. 


tion. 


ture. 




12.0 


276 


72 


39-1° 


Temperature of room 16° C. 


12.5 






... 


0"48 c.c. 1 in 500 tutin solution injected under skin of back. 


12.16 


192 


132 




Slight dyspnoea. Ear vessels somewhat dilated. Animal has been 
resting since injection. 


12.23 


180 


144 






12.28 


192 


168 




Lay down and stretched hind legs back. 


12.35 


174 


156 


37 -6* 


When lifted up the hind feet clung to the fingers. 


12.54 


198 


132 




Slight trembling occurred at intervals, and terminated in a 
convulsion two minutes later, which was mainly tonic, and 
seemed to prove almost fatal. 


12.58 








Tremors of mouth, and later of fore limbs, recommenced and 
gradually increased in intensity. The convulsion lasted two 
minutes. Afterwards the animal lay on venter with hind legs 
stretched backwards as before. 


1.7 


240 


84 


371° 


Fine tremors present apparently over whole body. 


1.20 


236 


144 




Tremors more marked ; otherwise same. 


1.35 


216 


108 


37-6° 




2.8 


276 


108 


38-r 


Somewhat better. 


3.8 


276 


108 


387° 


Much better ; has not moved since last note. 


4.5 


258 


84 


39-2° 


Apparently well. 



Experiment II. the effect of a minimal convulsant dose. Experiment III. shows the 
result of administering a non-convulsive dose. P refers to the number of heart-beats, 
and R to the number of respirations per minute ; T refers to rectal temperature. 

Experiment III. — 0-0005 grm. Tutin per kg. Rabbit. Weight, 1100 grm. 



Time. 


Heart- 
beats. 


Respira- 
tion. 


Tempera 
ture. 


Observations. 


12.0 


252 


66 


395° 


Temperature of room 16° C. 


12.4 








- 28 c.c. 1 in 500 tutin solution inje-ted under skin of back. 


12.10 


252 






Animal moving about. 


12.15 


216 


156 




Appears slightly depressed. Ear vessels dilated somewhat. 


12.21 


192 


180 




Dyspnoea Heart-beats feel more forcible. 


12.28 


186 


228 




Seems distinctly depressed, and shows no inclination to move. 


12.33 


156 


228 


37-6° 


When animal was lifted up the movements of the hind legs were 
found to be weaker and inco-ordinate. 


12.44 


168 


144 


. . . 


Trembling slightly. Ear vessels more contracted. 


12.53 


156 


156 






1.3 


168 


108 


38-0° 




1.18 


204 


144 






1.33 


228 


96 


387° 


Trembling still present. 


2.0 


264 


84 


39-6° 


Apparently well. 


3.3 


252 


72 







294 PROFESSOR C. R. MARSHALL ON 

To a rabbit weighing 700 grm. a dose of tutin corresponding to 00002 grm. per kg. 
body-weight was given without any effect beyond a slight one on the pulse, respiration, 
and temperature. The pulse fell from 228, the rate before the injection, to 192 thirty 
minutes after the injection ; the respiration rose from 54 to 60 ; the temperature fell 
in the hour succeeding the administration from 39 "0° C. to 38 '2° C. 

Three days afterwards a dose corresponding to 0"0003 grm. per kg. body -weight was 
given to the same rabbit. The heart-beats fell from 276 to 180 forty minutes after 
the injection, and the temperature from 39 "3° to 37*6° one hour after the administra- 
tion. The frequency of the respirations rose from 60 to 120 in thirty minutes. 

On Guinea-pigs. — Similar symptoms to those described as occurring in rabbits 
occurred in guinea-pigs. These animals appear, however, to be more susceptible than 
rabbits to tutin, as the following experiment shows :— 

Experiment IV. — - 001 grm. Tutin per kg. Guinea-pig. Weight, 458 grm. 



Time. 



Observations. 



12.35 - 46 c.c. 1 in 1000 tutin solution injected under skin of abdomen. 

1.14 Commenced to jerk head. After this ceased, animal began to move about carefully, the 
feet appearing to cling to bench. 

1.28 Coarse tremors of head, which extended to whole body. 

1.32J Convulsion commencing in fore part of body. The animal then rolled over several times, 
and finally stood up on toes and remained almost in a tonic spasm. The convulsion 
lasted 90". Afterwards the animal moved about carefully and somewhat stiffly. 

1.45 Convulsion similar to previous one; lasted one minute. 

1.48 Clonic movements; these caused the animal to run about the bench. They were followed 

by tonic contraction of the head muscles — eyes closed, ears drawn back, etc. — and by 
further clonic movements of limbs. On taking up the animal it entered into a tonic 
convulsion which lasted about a minute, and then the animal died — there was sudden 
and complete relaxation of the muscles, but the heart beat slowly for nearly a minute. 



Frogs. — In frogs also the most prominent symptoms are tonic and clonic 
convulsions. After a short period, during which the animal is apparently normal, it 
appears to be depressed and weaker, and gradually sinks down on the bench. The 
respiratory movements are usually deeper. If a large dose has been given, repeated 
opisthotonic convulsions, sometimes accompanied by a cry apparently resulting from 
spasm of the laryngeal muscles, follow. The convulsions may come on suddenly or 
may commence in retraction of the head, or, more rarely, in spasm of the hind legs. 
In the intervals, if these are short, the animal lies exhausted, but if of long duration it 
may move inco-ordinately and assume bizarre positions. Usually exhaustion supervenes 
and the animal appears dead. 

After smaller doses the convulsions tend towards a clonic type, although with 
freshly caught frogs, transient opisthotonos, more rarely emprosthotonos, is common 
even with small doses. In frogs kept in confinement for some time, the convulsions 
develop in a more definite manner- — the eyes are closed, the head retracted, then the 
fore limbs and lastly the whole body (opisthotonos) are affected. 



THE PHARMACOLOGICAL ACTION OF TUTU. 



295 



Experiment V. — 0-066 mg. Tutin per grm. Frog (R. temporaria — fresh caught). Weight, 15 grm. 



Time. 



11.43 
11.55 
11.56 
11.57 
12.0 



12.5 

12.10 

12.11 

12.12 

12.13 

12.15 
12.16 
12.18 

12.20 
12.25 
12.35 
12.40 

1.0 

2.0 
3.0 



Observations. 



1 c.c. 1 in 1000 tutin solution injected into dorsal lymph-sac. 

Sunk on to venter ; creeping about. 

Assumed bizarre position, hind legs being stretched out at right angles to body. 

Very transient opisthotonos. 

Blowing on the animal produced transient opisthotonos. Afterwards the frog tried to walk, 
but was unable to raise itself from the ground, hence it merely slid along on its venter. 
Hind-leg movements very stiff. 

Still sliding along ground for short distances. 

Made a few short jumps (clonic convulsion) ending in a slight opisthotonic attack. 

Another similar attack. 

Severe clonic convulsion (back arched) with cries. Belly appears blown out. 

Repeated clonic convulsions without cries. The frog remained on its back after the attacks 
ceased. 

Slight movements — mainly attempts to turn over — -with mild cries. 

Placed on venter. At once assumed a peculiar position closely resembling emprosthotonos. 

Made few attempts to spring, croaking each time. Whole body presents a blown-out 
appearance. 

Moving about, but movements incoordinate. Croaking. Eyes opened; corneal reflex sluggish. 

Placed on back. Remained motionless at first, then made unsuccessful attempts to turn over. 

Has not succeeded in turning over. Placed on venter. 

Moving about a little, with nose on ground. Movements incoordinate and apparently ex- 
hausting. Lymph-sacs blown out. 

Has been quiet since last note ; now made repeated futile attempts to move. Presents a less 
blown-out appearance. Placed on back. 

Quiet. Has made a few attempts to turn over. Placed on venter. 

No distinct improvement. Has moved inco-ordinately at times. Pithed. All stiffness im- 
mediately disappeared. Heart beating well. 



Trout and Tadpoles. — Trout, a few weeks old, when put into 1 in 1000 tutin solution, 
showed tremors of the tail in eleven minutes, marked twitches in thirty minutes, and 
tetanic convulsions in less than an hour. Tadpoles were affected similarly. 



Action on Medullary Centres. 

The protocols given on page 293 show that in rabbits tutin diminishes the rate 
of the heart-beat and increases the frequency of the ^respiration, and that the two effects 
are in each case nearly reciprocally proportional. This relative effect is seen better 
when the experiments are plotted as graphs (figs. 1 and 2). Fig. 1 shows the action of 
a minimal effective dose of tutin, and demonstrates the slightly irregular reciprocity in 
the effect on the pulse and respiration which is invariably seen after such small doses. 
After moderate doses of tutin the curves of frequency of respiration and pulse become more 
uniformly reciprocal — in the graph shown (fig. 2, Experiment II.) each curve is almost 
a mirror curve of the other. This greater uniformity in the case of moderate doses is 
probably due to the greater stimulus assuming more effective control of the medullary 
centres, thus rendering them less liable to accidental variation from without. That this 
influence on the heart-beat and respiration is due to an action on the medullary centres 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART II. (NO. 13). 45 



296 



PROFESSOR C. R. MARSHALL ON 



is suggested by the curves, and is supported by the fact that division of the vagi 
annuls the action of tutin on the circulation but not on the respiration. Systematic 
observations of the respiration and pulse have not been made after the administration 
of large doses mainly owing to the fear of exciting convulsions by taking the pulse ; 
but from the occasional notes made, it would seem that the increase in the frequency of 
the respiration and the diminution in the rate of the heart-beats are not proportionately, 
and may not be actually, as great after large as after relatively small doses. 

Heart Resp. 

120 



110 



100 



90 



80 



70 



60 













































































300 




























































































i 








































































1 


\ 

i 








































































i 


\ 










































































\ 




























































i 


> 












\ 






,1* 












































,< 


1 
























1 




/ 
























































\ 














\ 




/ 


\ 






















































\ 














\ 


i 
I 




\ 


























< 


1' 




































l 




\ 


I 




\ 


L 














































250 
















1 




\ 


i 






\ 




















































\ 










1 












V 
\ 






















































\ 








1 












\ 






















































\ 








1 














[ 










































































\ 

\ 










/ 


1 








































< 


' 
i 






► 
















\ 






J 


' 












































\ 




I 


















\ 






/ 














































\ 




1 


















\ 




/ 


I 
















































\ 


1 




















\ , 


/ 


















































\ 


/ 


















/ 


A 






































200 














\ 










/ 


s 


•* 


/ 


/ 






~~ 


— * 


^^ 














































\ 










/ 




s 


/ 
















■^ 


^ 
^ 






































/ 


, ! 


> — 


\ 






1 




























~--l 


> 






























/ 


/ 






\ 


\ 


1 
























































. i 


| 


— t 


/ 

i — 










\ 


I 


























































\ 


/ 

































































































































































































































































































Minutes. 



20 



40 



60 



80 



100 



120 



140 



160 



Fig. 1. — Diagram showing effect of 0"3 mg. tutin per kg. body-weight on the frequency of the respiration and the heart in a 
rabbit (weight, 750 grm.). Continued line = number of heart-beats per minute. Broken line — number of respirations 
per minute. 

During deep anaesthesia the reciprocal relationship between the frequency of the 
respirations and the heart-beats after tutin is lost. In this condition the administration 
of tutin usually diminishes the number of respirations per minute, while the circulation 
remains unaffected or is affected to a relatively slight degree. With chloroform as the 
anaesthetic, slowing of the respiration after the injection of tutin was invariable. When 
ether was employed, increased rapidity of the respiration followed if the anaesthesia was 
light, and slowing if the anaesthesia was deep. The following protocol shows the effect of 
tutin on the blood pressure, pulse rate, and respiration during chloroform anaesthesia : — 



THE PHARMACOLOGICAL ACTION OF TUTU. 



297 



Experiment VI. 



-Rabbit. Chloroform. Blood Pressure from Right Carotid Artery. 
Respiration recorded by Head's Method. 



Time. 


Blood Pressure 
in mm. 


Pulse Rate 
per 10". 


Respirations 
per 10". 




Remarks. 




3.34 


61 


27 


14-3 








3.35 








000025 grm. 


tutin injected into left ext. 


jugular vein. 


3.40 


61 


245 


13 7 








3.45 


61 


24 


13-8 








3.50 


62 


23 


13 








3.53 


65 


22-5 


11-5 








3.55 


66 


23 


8 








3.57 


65 


23-2 


10-8 








4.5 


66 


22 


16 








4.7 








000025 grm. 


tutin injected. 




4.12 


65 


22 


13-5 








4.17 


65 


21 


13 








4.22 


65 


21 


12-3 








4.25 


63 


21 


123 








4.27 


57 


20 


107 








4.30 


52 


20 


13 









Subsequently, a small dose of coriamyrtin (0"0001 grm.), and afterwards two doses of tutin (0-0005 grm. 
and - 001 grm.), were injected. In each case the blood pressure and pulse rate fell slightly, and the fre- 
quency of respiration very distinctly. 



Heart 



250 



200 



150 













































































































< 


1 




















































\ 




















































\ 




















































\ 




















































\ 


v 




















1 


► 






























\ 


















,' 




\ 
\ 






























\ 














-- 


' 




































\ 










/ 


p 














\ 




































• 


















\ 






























• 


' 




















^ 


» - 






-H 


► 




















1 






































































































\l 




















































' \ 


















































/ 




















































/ 




\ 






















/ 


:>-_ 






















/ 


) 




\ 






















/ 








^< 


> 














/ 






o 


\ 


















/ 


























/ 








\ 


\ 














/ 


























/ 












\ 


\ 












/ 


























/ 
























/ 


























< 


1 






















c 


> 









































































































































































































































Resp. 



150 



100 



50 



Minutes 10 20 30 40 50 

Fig. 2. — Similar diagram showing effect of 1 mg. tutin per kg. body-weight up to occurrence of convulsion. Rabbit, 950 grm. 



298 



PROFESSOR C. R. MARSHALL ON 



The slowing of the respiration produced by tutin during anaesthesia is not altered 
by division of the vagi. In a rabbit anaesthetised with chloroform the injection of 
0*0005 grm. tutin intravenously before division of the vagi reduced the number of 
respirations from 115 to 60 per minute; after section of the vagi the same dose 
caused the number of respirations to fall from 85 to 50 per minute. 

If the brain is excised and the administration of the anaesthetic afterwards stopped, 
the injection of tutin produces an effect similar to that observed in an uninjured 
animal. The action of a large dose up to the onset of convulsions is seen in the follow- 
ing experiment : — 

Experiment VII. — Rabbit. Ether. Section through Posterior Part of Mid-brain. Brain above Level 
of Section removed. Amesthetic stopped. Blood Pressure from Right Carotid Artery. Respirations 
counted. 



1 

Time. 


Blood Pressure 


Pulse Rate per 10". 


Respirations 
per 10". 




in mm. 


3.30 


65 


44 


9 


3.35 


0"01 grm. tutin injected 


into right facial vein. 




3.36 


71 


44 




3.37 


83 


38 


11 


3.38 


88 


34 


12-5 


3.40 


93 


36 


155 


3.42 


96 


36 


17 


3.43£ 


Clonic convulsions. 







The effect, particularly on the pulse rate, is much less than in a normal animal, but 
this is probably due to the medullary centres being still influenced by the previous 
administration of the anaesthetic. 

In a normal animal the dyspnoea produced by tutin is largely inspiratory in 
character, and in the slow respiration occurring during anaesthesia it is the inspiration 
which is mainly affected. It is not only increased in depth, but, when completed, is 
also maintained for a relatively long period. The effect is shown in fig. 3, which also 
illustrates the action of an allied substance, coriamyrtin, injected subsequently. The 
latter tracing further illustrates the irregularity of the respiration which not 
infrequently occurs after the administration of these substances. The respiration in 
this experiment and in most others, owing to the late appearance of the effects, was 
recorded by the diaphragm method described by Head ; but the same result was 
obtained when a record of the respired air was taken by connecting the tracheal 
cannula with a tambour. 

The slow breathing produced by moderate doses of tutin or coriamyrtin can 
usually be abolished by pushing the anaesthetic, but if a large dose has been ad- 
ministered deeper anaesthesia causes further slowing, which is quickly succeeded by 
cessation of the respiration. If constant anaesthesia is maintained, the respiration after 
a variable period generally assumes its previous frequency ; but a certain degree of 



THE PHARMACOLOGICAL ACTION OF TUTU. 



299 







a .£ 






o 



&>^ a 
.2 3 o 

° * » 

S - +> 

^ S^5 



OKI , 

n M 3" 

13 co 

a/ 60 






CD 

'5 S"is° 







O « 

. CO 

^-* CD 

rQ CD O 

(3 S 3 



■■gaj 
i l 1 

*-" C3 k 

*X3 co cd 
C .a 

a - eh 

O co 



co Tj 

" CI 



S [I 



.3 3 II -2 



«2 


a. 


3! 




Sn 


CO 


H 


Ph 


td 


CD 


Fh 


CO 


| 


II 


« 






£ 


V 


CO 


S*3 


O 




o 
















£ 









300 PROFESSOR C. R. MARSHALL ON 

slowing may reappear for short intervals independently of any obvious convulsive 
action (fig. 3). Later, if the dose has been sufficient, the respiration becomes irregular 
in extent, although it usually retains a fairly regular rhythm. This condition may 
continue for a long period, but finally it leads to deficient aeration and to cessation of the 
respiration. Variations in the respiratory tracings have occurred. In two instances a 
regular alternation of shallow with slower deep respirations developed ; in other cases 
the respirations during the absence of convulsions were characterised chiefly by 
prolonged inspirations. The influence of the convulsions on the respiration during 
anaesthesia was relatively slight. Sometimes a diminution in the frequency of the 
respirations accompanied the convulsion, but generally no distinct change in the 
respiratory rhythm occurred. The convulsions in these experiments were purely 
clonic in character. 

During anaesthesia the circulation also is very little influenced by the administration 
of moderate doses of tutin, or even by the convulsions produced. When chloroform 
was employed as the anaesthetic a slight rise of blood pressure and slowing of the 
heart's frequency followed the injection of the drug and coincided with the effect on 
the respiration, but not infrequently neither blood pressure nor heart rate was 
changed. With ether as anaesthetic a rise of blood pressure and a diminution in the 
frequency of the heart were more common, and were greater in extent than with 
chloroform. The difference in all probability is mainly due to the lighter degree of 
anaesthesia produced in rabbits and cats by ether. Of the three chief medullary 
centres, the vasomotor appears to be least susceptible and the respiratory the most 
susceptible to tutin injected during anaesthesia. 

The Convulsions. 

The convulsions produced by tutin are typically epileptiform in character. After 
a non-lethal but convulsant dose the earliest manifestations of the convulsant action 
of the drug are rapid head-noddings, or tremors of other parts of the body, and an 
increase in tone of the muscles, and especially of the limb muscles. These tremors 
are followed by clonic movements terminating in a tonic spasm. The clonic spasms 
commence, in the early convulsions generally, and in the later convulsions always, 
in twitches about the face and head — in movements of the mouth, or winking of 
the eyes, or twitching of the ears — which may at first be unilateral. There may be 
also rotation of the head to one or other side. Then follow clonic movements of the 
fore limbs and later of the hind limbs, and the termination of the clonic spasms in a 
tonic convulsion, which may or may not be followed by clonic movements. In some 
fits clonic movements only occur. After large doses a tonic fit may start the series 
of convulsions, but if the animal recovers from the early fits clonic movements always 
form part of the convulsion. Still later the animal loses consciousness, and the 
periodic fits become replaced by continuous clonic movements of the limbs, which only 



THE PHARMACOLOGICAL ACTION OF TUTU. 301 

terminate with the death of the animal. This may occur from exhaustion, but as 
the movements may apparently continue for long periods (an hour or more) the animal 
in these cases was generally killed by chloroform. 

In the clonic movements occurring in rabbits, both fore limbs and both hind limbs 
act in concert. The two hind limbs also act symmetrically, flexion and extension 
movements occurring in the two limbs at the same time. The fore limbs may act 
symmetrically or alternately. Clonic movements of one limb in a normal animal 
I do not remember to have seen. In anaesthetised animals, however, clonic move- 
ments of a single limb, and even independent movement of all the limbs, are not 
uncommon. 

The development of the convulsions in a decerebrate animal after a moderate dose 
of tutin is seen in figs. 4-6. The animal (rabbit, weight 1400 grm.) was anaesthetised 




Fig. 4. — Initial contractions of left fore foot in decerebrate rabbit after tutin. 
Rabbit, 1400 grm. Ether. 3.25, section just in front of pons; brain 
above removed ; ether stopped. 3. 55, 2 mg. tutin intraperitoneally. 
4.3, twitches commenced in all limbs. Time in seconds, x | linear. 

with ether. At 3.25 the brain was cut through at the junction of the mid-brain and 
pons, and the part above the level of the section excised. The administration of ether 
was then discontinued. Hooks were put into each fore foot and the right hind foot, 
and were connected by cords working over pulleys to weighted writing levers in such 
a manner that the levers trebled the extent of the limb movements. It was found 
impossible to maintain the two fore limbs in the same relative position to cords and 
pulleys, and this is the explanation of the complete want of uniformity in the tracing 
shown in fig. 6. At 3.55, 0*002 grm. tutin (1 c.c. 1 in 500 solution) was injected 
intraperitoneally. Eight minutes later (4.3), convulsive twitches appeared in all the 
limbs at once. At first the twitches were single (fig. 4), but later they showed a 
tendency to doubling, and from this state developed into short tonic contractions 
lasting one and a half to two seconds with clonic contractions at their height. These 
occurred every minute for the next few minutes, and were followed by an interval of 



302 PROFESSOR C. R. MARSHALL ON 

one and a half minutes terminated by a few irregular clonic contractions and a longer 
interval of rest (nine minutes). At 4.21, convulsions similar to those shown in fig. 5 
commenced, and continued with great regularity for three and a quarter minutes. (A 
second dose of tutin, 0*001 grm., was injected at 4.23^.) The convulsions then gradually 
became longer owing to the development of an increasing terminal clonus until, at 4.28, 




4.22J 

Fig. 5. — Later contractions of right fore limb in decerebrate rabbit after tutin. Same animal, x j linear. 

they became similar to those shown at the beginning of fig. 6. They were now 
continuous, a tonic stage with superposed slight clonic contractions of about three 
seconds' duration alternating with a clonic stage lasting about fifteen seconds. As this 
state seemed likely to continue, the effect of ether was tried, and it will be convenient 
to describe it here. 

Influence of Anesthetics on the Convulsions. 

The effect of inhaling ether on the convulsions of the decerebrate animal is shown 
in fig. 6. At the point marked by the arrow the animal was made to breathe ether 
vapour of the strength used for inducing anaesthesia for operations, and this was 
continued for eighty seconds. The convulsions ceased almost immediately after the 
commencement of the inhalation, and very slight contractions recommenced two minutes 
after its discontinuance. From this point the contractions, which at first were regular 
in rhythm, increased in intensity, and finally assumed the alternating tonic-clonic 



THE PHARMACOLOGICAL ACTION OF TUTU. 



303 



character seen at the commencement of the tracing twenty-two minutes after stopping 
the ether inhalation. Subsequently the effect of a very small concentration of ether 
vapour was tried. The anaesthetising tube was merely placed on one of the mouths 
of the Woulff's bottle, and a second limb of the tracheal cannula was left open. The 
effect, although less marked than in the preceding case, was strikingly sudden. Only 
one tonic-clonic convulsion occurred after the commencement of the inhalation. Within 




4.42 



Fig. 6. — Effect of inhalation of ether on tutin convulsions in decerebrate rabbit. Same animal as preceding 
figures. Anaesthetising concentration of ether vapour given at point marked by arrow. R = right, L — left 
fore limb. Time in seconds, x § linear. 

twenty-three seconds true convulsive movements had ceased and become replaced by 
single contractions, occurring at first in small groups, and later alone. These con- 
tractions ceased eighty seconds after the commencement of the inhalation, which 
was continued for two and a quarter minutes. JRegular small contractions commenced 
ninety seconds later, and the typical tonic-clonic convulsions quickly developed 
and became distinct in one minute. A second similar inhalation of ether produced 
practically the same result. 

It is evident, therefore, that the convulsions occurring in decerebrate animals are 
TRANS. ROY. SOC. EDIN, VOL. XL VII. PART II. (NO. 13). 46 



304 PROFESSOR C. R. MARSHALL ON 

very susceptible to the action of ether. When the cerebrum is intact, ether appears 
to be less active in arresting convulsions. Even with fairly deep anaesthesia, clonic 
convulsions are produced by relatively small doses of tutin ; and after large doses of 
tutin the convulsions do not cease when the concentration of ether inhaled is increased, 
until shortly before the respiration is paralysed. As will be shown later, when one 
cerebral hemisphere is excised the injection of moderate doses of tutin causes 
convulsions only on the same side ; and since the mid-brain and pons are still intact 
in this experiment, it would seem as if the ether had prevented the development of 
the convulsions from this part of the brain. 

The effect of diminishing the quantity of ether administered has been observed in 
experiments in which the contractions of individual muscles were being recorded. 
To take an example. With a certain concentration of ether vapour the contractions 
of the Tibialis Anticus were fairly uniform, and registered 3 mm. in height. After 
reducing the concentration of ether vapour about 40 per cent., the contractions assumed 
a more convulsive type in twenty-seven seconds, and many at thirty seconds registered 
27 mm., and at fifty seconds 35-40 mm. in height. The weaker ether inhala- 
tion was continued for sixty-six seconds, when the concentration was raised to its 
previous level. The contractions commenced to diminish in eighteen seconds, and 
in forty seconds they had almost reached their previous form. This occurred twelve 
seconds later. 

The effect of chloroform on the convulsions was not graphically recorded, except 
indirectly in tracings of the respiration by Head's method ; but from these and other 
experiments it appeared to be more powerful, for apparently the same degree of 
anaesthesia, than ether in arresting the convulsions. 

Muscles Affected by the Convulsions. 

A few experiments were made to determine the way in which various individual 
muscles are affected by the administration of tutin. They were undertaken primarily 
with the object of determining the cause of the bizarre positions and curious movements 
which sometimes occur after tutin. In the experiments the contractions of the 
following muscles were registered : — Digastric, Mylohyoid, muscles of pinna of ear, 
Biceps, Triceps, Flexor Carpi Radialis, Flexor Sublimis Digitorum, Extensor Communis 
Digitorum, Rectus Femoris, Gracilis, Tibialis Anticus, Extensor Longus Digitorum, 
Flexor Longus Hallucis, Plantaris, and Diaphragm. The muscles of the shoulder 
and hip were not isolated, but their movements were registered in several experiments 
by more distally situated muscles which had been isolated. The movements of the 
muscles of the abdominal wall and the prevertebral muscles were observed but not 
registered. So far as could be determined, all the voluntary muscles of the body are 
influenced by tutin, although not simultaneously. The muscles were isolated from 
the surrounding tissues as far as possible without disturbing their blood and nerve 



THE PHARMACOLOGICAL ACTION OF TUTU. 305 

supply, and their distal tendons were connected by hook or ligature to cords working 
over pulleys and ending in weighted levers. All these experiments, with the exception 
of those on the diaphragm, which were made on rabbits, were made on cats under the 
influence of ether. 

The records of the different convulsions are far from uniform : no convulsion either 
in the same or in different animals was exactly similar to another. After small doses, 
and generally after large doses, the early convulsions commenced, as previously 
described, in the muscles of the head and neck, and affected the fore limb muscles 
before those of the hind limbs ; but in later convulsions great irregularities appeared, 
and the posterior limb muscles were not infrequently affected first. Even muscles of 
the same limb serving similar functions appeared at times to be variously affected, and 
the form of the convulsions in different muscles was not infrequently diverse. Some 
of these points are shown in figs. 7 and 8. Fig. 7 illustrates in a marked degree the 
usual condition of a convulsion commencing in the muscle of the anterior limb before 
that of the posterior limb, and it also shows, what is frequently the case, the greater 
tonus in the muscle of the hind limb. The cat received an injection into the peritoneal 
cavity at 3.25. Twitches occurred in the Flexor Carpi Radialis at 3.49 and 3.51, and 
short tonic contractions in the Tibialis Anticus with occasional contractions in the 
first-named muscle at 3.55. At 4.5, short convulsions, clonic and tonic-clonic in 
character, and lasting six to eighteen seconds with intervals of three to five 
seconds, commenced in both muscles and continued until 4.16. A longer interval 
of rest then occurred, followed by a convulsion similar to that shown. This was 
succeeded by a rest of fifty-five seconds, which in turn was followed by the 
convulsion illustrated. 

Fig. 8 shows a convulsion commencing in the Plantaris earlier than in the Extensor 
Longus Digitorum of the same limb or the Flexor Sublimis Digitorum of the same 
side, and it also serves to illustrate the variations in form of the convulsions in different 
muscles. 

The experiments seem to indicate a marked loss of co-ordination in the motor areas, 
and they explain in some measure the irregular movements and want of control seen in 
the earlier stages of severe tutin poisoning, as well as the irregular clonus of the 
late stages. 

In an experiment in which the contractions of antagonistic muscles — viz. Biceps, 
Triceps, Gracilis, and Eectus Femoris — were registered, inhibition of the extensor 
muscles occurred during the early twitches of the flexors ; but later, when more prolonged 
tonic contractions and convulsions developed, no distinct inhibition was present. 

The muscles of the trunk, shoulder, and hip appeared from direct observation to be more 
easily affected by tutin than the more distal muscles of the limbs, but the diaphragm 
seemed less easily influenced. The irregularity of the contractions of the diaphragm 
seen in many of my tracings was undoubtedly caused mainly by contractions of the 
muscles of the abdominal walls and the prevertebral muscles, but in some instances 



306 



PROFESSOR C. R. MARSHALL ON 




w 



o 



u 

p 


o 




o 


CO 


a> 


p 


cfi 


o 


.3 


e 


OJ 


<j 


s 






GO 


H 






C9 








3 








H 




T3 





03 



# 



C=H 



3 

a 

S3 






e~ 



THE PHARMACOLOGICAL ACTION OF TUTU. 



307 



the diaphragm also appeared to 
take part in the clonic fit. No 
prolonged tonic contraction of 
the diaphragm was obtained or 
observed. 

Seat of Origin of the 
Convulsions. 

The spinal cord is slightly, 
if at all, affected by tutin. In 
frogs pithed through the 
occipito-atlantal membrane, the 
injection of tutin produced no 
convulsions ; and if the tutin 
was injected and the con- 
vulsions allowed to develop 
before pithing, this operation 
caused the convulsions to 
cease, and they did not recur. 
In rabbits anaesthetised with 
ether or chloroform, or with 
the brain above the quadrate 
bodies excised, section of the 
spinal cord previous to the 
injection of tutin prevented 
the occurrence of convulsions or 
of increased reflex movements 
below the point of section ; and 
section of the spinal cord after 
the development of convulsions 
caused complete cessation of 
these below the point of section, 
and they did not recur.* 

* Gottlieb (Arch. f. exp. Path. u. 
Pharm. vol. xxx. p. 21) has shown that in 
dogs kept for several days to allow the 
shock of the operation to pass away, con- 
vulsive movements can be produced in the 
hind limbs by picrotoxin after section of 

the spinal cord, and that in certain fishes and amphibians and very young mammals convulsive movements can be 
induced by this drug behind a spinal transection soon after the operation. He failed, however, to get any evidence 
of spinal stimulation in pithed frogs. See also note at the end of this paper. 




308 PROFESSOR C. R. MARSHALL ON 

In frogs pithed anteriorly to the optic lobes, convulsions similar to those occurring 
in a normal frog were produced by tutin, but after destruction of these lobes muscular 
stiffness only resulted. 

In rabbits and cats which had the brain above the pons removed convulsions were 
still produced by tutin (figs. 4-6) ; and when the brain was divided just behind 
the pons, muscular stiffness and increased reflex excitability were produced by its 
administration, but less readily than convulsions with the pons intact. With the 
quadrate bodies in place, convulsive movements seemed to be somewhat more easily 
induced than when the section was made immediately in front of the pons. One 
cannot, however, be dogmatic on this point. In an experiment (rabbit, 1575 grm.) in 
which the brain was cut through in front of the corpora quadrigemina, and the 
anaesthetic subsequently stopped, the convulsions induced by 10 mg. tutin intraperi- 
toneally were diminished but not abolished by section between the anterior and 
posterior bodies (the hind limbs still twitched), but all limb movements ceased after 
section behind the posterior bodies. There can be little doubt that shock played an 
important part in producing this result ; but nevertheless it gives some support to the 
view expressed that the presence of the mid-brain increases to some extent the 
susceptibility to tutin. 

Notwithstanding the susceptibility of the lower centres of the brain to this substance, 
the cortical centres are probably mainly affected under ordinary circumstances. This 
is suggested by the following experiment. A cat (1750 grm.) was anaesthetised with 
chloroform and the anaesthesia maintained with ether, and at 3.30 the left cerebral 
hemisphere was removed. The fore limbs were connected by cords working over 
pulleys to weighted levers which registered their movements. A.t 4.0, 0"001 grm. tutin, 
and at 4. 1 6, 0'0005 grm. tutin, were injected intravenously. Slight movements of the left 
fore foot occurred at 4.19; and at 4.20 short tonic contractions with superposed 
clonus developed. These continued at irregular brief intervals for seventy seconds, 
when a well-marked and prolonged tonic-clonic convulsion occurred, which gave place 
to the convulsions shown in fig. 9. These continued until 4.29 ; afterwards, intervals 
of rest, at first of fifteen seconds and later of longer duration, separated the convulsions, 
which also became more severe in type. The animal was finally killed with the anaes- 
thetic. The convulsions on the left side were identical in character with those observed 
in other experiments with the cerebrum intact. No convulsions and no movements 
occurred on the right side. Some slight movements were recorded from this side in 
the earlier convulsions (see fig. 9), but these were found to be transmitted from the 
left side. When the shoulders were steadied, no movement was recorded on the 
right side and no contractions were felt. In view of the fact that convulsions may 
readily be produced by tutin after removal of both cerebral hemispheres, the absence 
of convulsions on the left side in this experiment can only be explained by assuming 
a greater susceptibility on the part of the lower centres affected by tutin to anaesthetics. 
This susceptibility has been shown to exist (p. 302). 



THE PHARMACOLOGICAL ACTION OF TUTU. 



309 



The chief seats of origin of tutin convulsions in rabbits and cats would therefore 
appear to be the cortical centres and the pontine nuclei. The share which each of these 
centres takes in producing the convulsions in unanaesthetised and unmutilated animals 
is difficult to prove ; but from the nature of the anxiety of the preconvulsive stage 




4.27 



Fig. 9. — Effect of tutin on fore limbs after excision of left cerebral hemisphere. Cat, 
1750 grm. Ether. 3.30, left cerebral hemisphere removed. 4.0, 1 mg. tutin intra- 
venously. 4.16, 0'5 mg. tutin intravenously. The small contractions of right fore 
limb (R) were transmitted from left shoulder muscles. Time in seconds, x f linear. 

aud the character of the early convulsions I am inclined to believe that the earliest 
action is upon the cortical centres. The continued clonus of the late stages, which is 
associated with complete unconsciousness, is in all probability due to the activity of the 
lower centres. 



Action on the Higher Centres. 

Owing to the ease with which reflex phenomena are obtained after the administra- 
tion of tutin, it is impossible to say definitely, from the observation of rabbits, that 
tutin in small convulsant doses has an action on the sensory and other higher centres. 
Decided depression and even sleepiness are seen in these animals soon after the injection 
of such doses ; but, although I am of opinion that this condition is the result, in part if 
not wholly, of an action on the higher centres, I was unable to convince myself that it 
is not secondary to the changes in the respiratory and circulatory systems, or due to a 
desire to keep the muscles at rest as long as possible. It is noteworthy, however, that 
the same degree of depression was not noted after coriamyrtin, which acts more rapidly 
but otherwise produces the same kind of convulsions as tutin ; and it is not seen after 
strychnine when given in sufficient doses to cause convulsions. After the administra- 



310 PROFESSOR C. R. MARSHALL ON 

tion of a lethal or nearly lethal dose of tutin to rabbits, a comatose condition usually 
develops in the later stages of the intoxication, and is in all probability a direct effect 
of the poison on the higher centres. 

The most convincing evidence of the depressant action of toot poison comes from 
cases of accidental poisoning in men. In all the cases of recovery described there has 
been no remembrance of eating tutu berries or of the convulsions which followed, and 
in some patients normal memory is said not to have returned for many months. Dr 
Giles, in a letter to Professor Easterfield and Mr Aston, describes the case of a soldier 
who suffered from violent convulsions, which he successfully treated by chloroform 
inhalations. " The next morning the man was sore from the violent strain the muscles 
had undergone, but was otherwise quite recovered, with the single exception that his 
memory was totally obliterated. The past was a perfect blank to him : he did not 
know his own name, who he was, or where he came from. This condition gradually 
passed away, and in a day or two he was as well as ever." Another example cited by 
the same medical man is that of two men who, having lost their way in the bush, had 
eaten tutu berries. One died ; the other was found gesticulating and behaving like a 
maniac. When rescued, " his memory was said by the witnesses to be a blank, and it 
only returned gradually." He also states that the late Dr Carl Fischer, who had 
commenced to investigate the poison, " had been obliged to discontinue his experiments 
because his memory became so impaired that he found himself constantly forgetting his 
appointments." Somewhat similar symptoms are described by Christie * as the result 
of an experiment on himself. After taking 8 J grains of an extract (representing 
about 800 grains of leaf) in two doses, vomiting commenced about four hours after the 
first dose, and slight twitches were felt in the arms and legs two hours later. A 
convulsion which was noted two and a half hours later was not remembered by him, 
but half a grain of morphine had been administered to stop the vomiting one and a half 
hours before. After recovery, " sensation and discrimination were dulled," and a month 
later parsesthesia and numbness were felt in the fingers and toes. These last-named 
symptoms continued for a month. There appears to be little doubt, therefore, that 
tutin, even in relatively small doses, exerts a decidedly depressant effect upon the 
higher functions of the brain. 

Influence on Temperature. 

In my experiments with rabbits, tutin invariably caused a fall of the body 
temperature, and after non-convulsant or small convulsant doses the fall was in most 
cases roughly proportional to the dose administered. This is seen in the following 
graph, more especially in the curves plotted for 0"5 mg., 1 mg., and 1"5 mg. per kg. 
body- weight respectively. The remaining curve of 3 mg. per kg. body- weight shows 
a greater proportionate effect than would be expected of this dose, but this is due to the 

* hoc. cit. 



THE PHARMACOLOGICAL ACTION OF TUTU. 



311 



fact that a younger animal was used for this experiment than for the others represented 
in the figure. The temperature appears to be more easily affected by tutin in young 
than in adult or old animals. After lethal doses of tutin the temperature in the few 
cases I have investigated fell from the first, and showed no tendency to recover, notwith- 
standing the continued convulsions. 

A fall of temperature has been shown by Harnack # and his pupils to occur after 
the administration of most convulsants. Harnack t was inclined at first to the view 
that they acted by stimulating the inhibitory mechanism of the heat-regulating centre 
in the brain, but later, as the result of further investigations with sodium santoninate 
and santonate.J he came to the conclusion that the antipyretic action of these 



Temp. 






















i 




























































































































































j 








































































































V 


A 


> 
































t 


> \ 






























































T 




_^ 


..( 


1 




39 




< 


C* 


> 














1 
I 






















^ - 


** 
















1 


— * 






















\ 


■s " 


N 










1 




















,'' 














































\ 


% 




\ 


























,< 


r 










.< 


i-" 


■ • 
















11/ 




^.K 


\ 














\ 


s 




< 


K 
















, 


■'' 










. 


-" 


' 












_, 


— 


" 
















38 










\ 


> 








N 


-N 








j-- ' 


'' 








^.< 


» 












. 








































\ 










X. 


' -s 












* 










*- 


















































t 


>v^ 


'" 








"< 


>" 










,-< 


i-~* 






























































"•- 


^ 














** 


































7 


7 


.< 


) 




37 






























o- 








































































































































































































































































































36 


































— 










































































































































































































































































— 






































































































Mins. 



30 



60 



90 



120 



150 



180 



Fio. 10.— Influence of tutin on temperature of rabbits. I = 0'3 mg. per kg. body-weight ; II = - 5 mg. per kg. body-weight ; 
III = 1 mg. per kg. body-weight ; IV = 1'5 mg. per kg. body-weight, adininistered hypodermically. 

convulsants is due to loss of heat owing to dilatation of the cutaneous vessels. This 
dilatation is brought about in all probability by stimulation of the vaso-dilator 
mechanism in the medulla, He does not think it follows that all convulsants which 
cause a fall of the body temperature act in this way, but he withdraws his former 
explanation of their mode of action. More recently Hayashi§ has shown that 
convulsants temporarily reduce the fever caused by heat-puncture, and in consequence 
he concludes that the action of these substances is a depressant one on the heat- 
regulating centre. 

Part of the evidence brought forward by Harnack in support of dilatation of the 



* Archivf. exper. Path. u. PharmaL, xxxviii. p. 397 [1897] ; xl. p. 151 [18971. 
t Ibid, xxxviii. p. 421 ; xl. p. 152. + Ibid., xlv. p. 272 [1901]. 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART II. (NO. 13). 



§ Ibid., 1. p. 247 [1903]. 
47 



312 PROFESSOR C. R. MARSHALL ON 

vessels being the cause of the fall of temperature is the appearance of the vessels of the 
car. These were seen to be dilated when the temperature was falling, and contracted 
when the temperature was rising. My observations after administering tutin largely 
confirm those of Harnack with sodium santoninate on this point. Unfortunately my 
experiments were made before the publication of his paper, and 1 did not pay special 
attention to this connection, but, where noted, dilatation of the vessels of the ear 
occurred during the fall and contraction during the rise of temperature. In a few 
instances the ear vessels were noted to be more contracted before the rise of tempera- 
ture commenced. 

Effect of Decomposition Products. 

The products of hydrolysis of tutin possess no peculiar pharmacological action. 
Insufficient hydrolysis yields a product with a variable tutin action dependent upon the 
undecomposed tutin. 

After heating with excess of 0'2 per cent, hydrochloric acid for fifteen minutes on a 
water-bath, neutralising with sodium bicarbonate, evaporating, extracting with alcohol, 
and re-extracting the inspissated alcoholic extract with ether, a colourless syrup was 
obtained which gave the physiological action of tutin, and from which crystals of tutin 
separated later. After heating the hydrochloric acid solution on the water-bath for an 
hour and extracting in a similar manner, a colourless syrup was obtained which did not 
deposit crystals of tutin, but which caused tutin convulsions after administration. On 
heating a solution of tutin in 1 per cent, hydrochloric acid on the water-bath for three 
hours, neutralising and extracting as before, a reddish oil insoluble in water and a 
brownish substance soluble in water were obtained, but neither produced any 
physiological effect. The sugar produced by hydrolysis has not been identified. I can 
corroborate Easterfield and Aston's * statement that it does not form a crystalline 
osazone. In itself, however, the sugar is of little pharmacological importance. 

Dilute solutions of alkalies appear to decompose tutin much more quickly than 
dilute acids. A solution of tutin in 0*1 per cent, sodium hydroxide, and one in 0'05 
per cent, calcium hydroxide, after raising to the boiling-point, cooling, and neutralising, 
produced no pharmacological action in doses which should have given, if any active 
principle had been present, a marked result. 



Comparison with other Convulsants. 

The best-known of the substances comparable with tutin are picrotoxin, coriamyrtm, 
and essence of absinthe. Many others cause convulsions by acting on the brain, but 
most of them produce other actions not given by tutin. In so far as their convulsant 
action has been investigated it would seem not to differ materially from that of the 

substances mentioned. 

* Loc tit. 



THE PHARMACOLOGICAL ACTION OF TUTU. 



313 



The form and course of the convulsions produced by the first-named substances 
seem to be the same. Even somersaulting is a symptom common to all. The main 
differences are in the time of appearance of the convulsions and the duration of the 
intoxication. Absinthe causes an earlier appearance of convulsions, but after non-lethal 
doses a shorter intoxication than tutin ; and the same is true of coriamyrtin and, to a 
less extent, of picrotoxin. The two last-mentioned substances are more powerful than 
tutin, in that they produce convulsions with smaller doses. This is shown in the 
following comparative experiments on frogs. The frogs were of the same weight 
(15 grm.), and the substance, dissolved in 0'6 per cent, sodium chloride, was injected 
into the dorsal lymph-sac in each case. The figures given indicate the number of 
minutes between the injection and the onset of the first distinct convulsion. 



Dose. 


Coriamyrtin. 


Picrotoxin. 


Tutin. 


1 mg. 
- 5 mg. 
0-08 mg. 
004 mg. 


1 

3 

10 

40 


7 

19 
No convulsion. 


15 

38 
No convulsion. 
No convulsion. 



The substance I have compared most frequently with tutin is coriamyrtin. This 
principle was isolated by Riban # from a species of Coriaria (C. myrtifolia) which is 
common in the South of France and other coasts bordering the Mediterranean. The 
relative effects of the two substances will be seen in the following brief protocols 
of experiments on the same animals. One set gives the results of hypodermic 
and the other of intravenous administration. A period of fourteen days in the first 
case and five days in the second occurred between the two experiments. 

Rabbit, 2000 grm. 



Time 




Time 




after 


Remarks. 


after 


Remarks. 


Injection. 




Injection. 







- 75 mg. coriamyrtin hypodermicaliy. 





2 mg. tutin hypodermicaliy. 


4' 


Dyspnoea. Head tremors. 


14' 


1 )yspncea. 


5' 


Clinging to table. 


60' 


Had no convulsive symptoms beyond 


9' 


Tonic and clonic convulsion lasting 




head tremors. 




one minute. 


64' 


2 mg. tutin hypodermicaliy. 




No further convulsions occurred. 


651' 


Clonic convulsion. 




The animal rapidly recovered from 


90' 


Slight clonic convulsion. 




the intoxication. 


107' 


Clonic-tonic convulsion. 






108' 


Convulsive movements of head. 






110' 


Opisthotonos. 






125' 


Tonic clonic convulsions. 






135' 


Chloroformed. 



Recherches expe'rimentales sur le principe toxique du redoul, Paris, 1863. 



14 



PROFESSOR C. R. MARSHALL OK 



Rabbit, 950 grin. R = number of respirations per minute. T = rectal temperature. 



Time 




Time 




after 


Remarks. 


after 


Remarks. 


Injection. 




Injection. 







R = 60. T = 396°. 





R = 48. T = 395°. 




1'44 mg. tutin into marginal vein of 




0'24 mg. coriamyrtin into marginal vein 




ear. 




of ear. 


4' 


Breathing more rapidly. 


r 


Very rapid breathing. 


5' 


R = 108. 


H' 


R = 204. 


6' 


R = 168. Head-noddings apparently 


5' 


R = 204. Head-noddings. 




synchronous with respiration. Slip- 


7' 


Powerful clonic terminating in tonic 




ping down on venter. 




convulsion. Lasted one minute. 


8' 


R = 204. Laid on venter with hind 




Quickly recovered. Then R = 84. 




limbs partially out. 


12' 


R = 204. Some twitching of ears. 


25' 


Head-nodding less marked, but slight 


15' 


Slightclonic convulsion of head, followed 




general tremors. Afterwards assumed 




by early stage of somersaulting move- 




characteristic (exaggerated springing) 




ment. 




posture. This strained position was 


17' 


Clonic convulsion of head. 




maintained for one hour. 


19' 


Quiet. R = 96. 


90' 


R=160. 


25' 


R = 108. 


104' 


Moved very carefully for a few steps. 


45' 


R = 60. Not moved since last note. 




Hind-limb movements somewhat 




Apparently normal. 




inco-ordinate. 


60' 


R = 60. T = 380°. 


120' 


R = 108. T = 36-8°. Recovering. 







These experiments show that in similar physiological closes tutin takes longer to 
act and exerts a more prolonged effect than coriamyrtin, but apart from these 
differences the two substances produce almost identical symptoms. The difference in 
time of action is important in relation to the question of tolerance, and it explains in 
some measure why, on the one hand, a second injection of tutin, administered after a 
long interval, should so rapidly induce the convulsions which failed to appear as a 
result of the first injection ; and, on the other hand, why relatively large quantities of 
coriamyrtin, if administered in divided doses at short intervals, can be given without 
inducing convulsions. 1 hope to publish experiments on this point later. 

Summary. 



1. Tutin, the active principle of the toot plant, causes epileptoid convulsions in 
various classes of animals. After large doses to rabbits the earliest convulsions 
commence with a general tonic spasm which is followed by clonic movements. After 
small convulsant doses the couvulsions begin with clonic spasms of the muscles of the 
head. These extend to the fore and then to the hind part of the body, and may 
terminate in a tonic spasm. The later convulsions after larger doses also assume this 
form. After the initial convulsions following small doses, somersaulting movements 
are not uncommon. 

2. The convulsions are mainly of cortical and pontine origin. Convulsive move- 



THE PHARMACOLOGICAL ACTION OF TUTU. 315 

ments can be obtained when the brain is divided below the pons, but they are less 
characteristic and are more difficult to induce than when the pons is intact. In frogs 
the optic lobes seem to be the most important centres affected. Unequivocal convulsive 
movements were not obtained below a section of the spinal cord (see, however, note 
below). 

3. The convulsions arising in the pons are very susceptible to anaesthetics. This 
probably explains why convulsions are limited to one side after excision of one cerebral 
hemisphere during the continuance of the anaesthesia. 

4. In unansesthetised rabbits the two fore limbs act together, and the two hind 
limbs act in concert and synchronously during tutin convulsions ; in the anaesthetised 
animal the limb movements are often asynchronous. Individual muscles of the limbs 
often act vicariously, and this probably explains the irregular clonus which the move- 
ments of the limbs show. 

5. Preceding the onset of the convulsions the medullary centres are stimulated. 
This effect is also produced by non-convulsant doses. 

6. All doses which produce an obvious action cause in rabbits a fall of body 
temperature. 

7. When heated with dilute caustic alkali solutions, tutin rapidly decomposes, the 
product being pharmacologically inactive. Prolonged heating with dilute acids leads 
to the same result. 

8. The substance most closely allied to tutin, namely coriamyrtin, is a more 
powerful convulsant, and, for similar physiological doses, is more rapid and more 
transient in action than tutin. 



Note. — Since the preceding paper was accepted for publication a research on the Physiological Action 
of Tutin by Fitchett and Malcolm has appeared in the Quarterly Journal of Experimental Physiology 
(vol. ii. p. 335; Oct. 1909); an extended description of the research, with protocols, being given by 
Fitchett in the Transactions of the New Zealand Institute (vol. xli. p. 286). The effect of tutin on 
various classes of animals is described, and, in so far as our experiments are common, our results are in 
general agreement. On a few physiological points we are not in accord. Fitchett and Malcolm have 
apparently not obtained any evidence of stimulation of the cardio-inhibitory centre, but they appear to have 
investigated the matter only in animals under anaesthetics. As I have shown, no decided stimulation of this 
centre occurs during surgical ansesthesia. 

They also attribute to the motor area a smaller and to the spinal cord a larger share in the production of 
the convulsions than I have done. They state that after inducing convulsions in a cat under chloroform 
by a large dose of tutin injected hypodermically and removing the right cerebral hemisphere, "the con- 
vulsions continued just as before, affecting both sides equally," and after scooping out the left cerebral 
hemisphere "the convulsions seemed to be quite unaffected." "The spinal cord was then divided in the 
mid-dorsal region. The movements below the level of section were absent or abolished at first, but four 
minutes after the section clonic and tonic spasms of the hind limbs were observed, and these continued 
intermittently until death ensued, a quarter of an hour later." The continuance of the convulsions after 
removal of the cerebral hemispheres is explained by the fact that, according to the protocol of the experiment 
(No. 152 in Fitchett's paper, p. 356), the anaesthetic was stopped when the tutin was administered — twenty- 
two minutes before the appearance of any symptoms. The only difference in our results is the effect 
obtained behind a transection of the spinal cord. All my transection experiments were made on rabbits, 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART II. (NO. 13). 48 



316 THE PHARMACOLOGICAL ACTION OF TUTU. 

but since the appearance of Malcolm and Fitchett's paper I have repeated two of the experiments on cats. 
In both cases the animals were anaesthetised with chloroform and the anaesthesia maintained with ether. 
In one animal 0"02 grm. tutin was injected into the external jugular vein, and five minutes after the hind 
limbs were seen and felt to twitch the spinal cord was completely divided at the level of the 11th dorsal 
vertebra. The convulsions in the parts innervated beyond the section immediately ceased, and they did not 
recur. Apparent convulsive movements in the hind parts appeared, but, as in experiments on rabbits, they 
were synchronous with the clonic convulsions occurring in the fore part of the body, and by careful 
palpation were proved to arise in the muscles innervated anteriorly to the spinal section. The muscles of 
the hind limbs were palpated for forty minutes after the transection, but no twitch was felt in them. In 
the second experiment the mid-brain was cut across just above the pons, and the brain above the section 
removed. The administration of ether was then stopped. Twelve minutes later (3.28) O02 grm. tutin 
was injected intravenously. Twitches of the eyelids and ears occurred at 3.37, and slight convulsive 
movements of the limbs at 3.42. These quickly became more severe, and at 3.51 the cord was completely 
divided at the level of the last dorsal vertebra. Slight independent movements of the hind limbs occurred 
at 3.54, and became more marked and assumed a more convulsive type later. Pressure on one hind foot 
excited convulsions in both hind limbs. The convulsions, however, were not so characteristic as those in 
the fore part of the body. The animal died at 4.15. This experiment, therefore, confirms the observation 
of Malcolm and Fitchett that tutin acts on the spinal cord in cats ; and the first experiment seems to show 
that this effect, like that upon the lower cerebral centres, is very susceptible to anaesthetics. 



( 317 ) 



XIV. — On the Life-history of Hydrobius fuscipes, L. By Frank Balfour-Browne, 

M.A. (Oxon)., F.Z.S. (With Three Plates.) 

(MS. received July 6, 1909. Read July 12, 1909. Issued separately February 1, 1910.) 

In following out the life-histories of some of the water-beetles I have been struck by 
the remarkable differences in habits and in structure between the members of a family, 
for instance, between Hydro philus and Hydrobius, two fairly closely related genera of 
the Hydrophilidse. The description of the life-history of Hydrophilus as given by 
Miger, Lyonnet, and others will, only to a small extent, apply to Hydrobius, although 
the suggestions which I am able to make with regard to certain facts in the life-history 
of the latter may throw light upon somewhat similar facts in the life-history of the 
former. 

Hydrobius fuscipes, L., is a very common water-beetle found throughout Great Britain 
and Ireland in almost every stagnant piece of water where there is plenty of aquatic 
vegetation, the favourite habitat appearing to be ditches or ponds where there is 
abundance of grass growing in or floating out on the water. The beetle is small, 
measuring about 6 or 7 mm. in length, slightly elongate, of a shining blue-black or 
brown-black, although often in bright light showing a beautiful velvety-green lustre. 
The legs in the typical form are reddish brown, the tibiae being tipped with darker 
brown or black. 

There is a variety of this species which is either not very common in Britain or of 
which collectors have not recorded their captures. This is H. seneus of Solier (Ann. Soc. 
Ent. France, p. 314, 1834), and according to Rye is H. chalconotus of Leach (1814) 
and of Stephens (Mand., ii. 128, 1829) (vide Ent. Mo. Mag., vii. p. 36, 1870-71). 
Rye records having taken this variety at Barnes (Surrey), and Morley, C. (Coleoptera of 
Suffolk, 1899) records it from " chalk pits at Little Blakenham (Suffolk, E.)." There is 
a specimen in the collection of the late Mr Arthur Chitty which he took in 1887 at 
Cuckfield Park (Sussex, W.), and recently Mr Donisthorpe took a specimen at Tottenham 
(Middlesex) (Ent. Rec, xx., July 1908, p. 184), and he mentions that it has been 
taken also at Sheppey (Kent, E.) and Woking (Surrey). It is recorded from Upper 
Teesdale (Yorks, N.W.) (Vict. County History, 1907). Mr T. Stainforth tells me it is 
common on the coast near Hull (Yorks, S.E.). I took a single specimen at Cambridge in 
1904, and it seems to be not uncommon at "The Kinnegar," Holywood, Co. Down. 
This variety apparently differs only in colour from typical H. fuscipes, and is of a 
metallic blue or green on its upper surface, but the brightness of the colouring varies 
greatly in different specimens and fades considerably after death. 

Thomson (Ann. de la Soc. Ent. de France, 1883, p. cxxxi.) separated as a distinct 

TRANS. ROY. SOC. EDIN., VOL. XLVII. PART II. (NO. 14). 49 



318 



MR FRANK BALFOUR-BROWNE ON 



species, which he named H. picicrus, a form which is less elongate and more rotund 
than the typical H. fuscipes and has thicker and more pitchy tibise. The pubescence 
of the hind femora also is not quite so extensive, and the punctuation of which it is the 
accompaniment is not quite so dense and fine (Sharp, D., Ent, Mo. Mag., xxi., 
84, 5, 1884-85). The elytral striae are more strongly punctured and therefore appear 
to be rather deeper than in H. fvscipes, and the punctuation of the upper surface is, as 
a rule, more coarse. The penis is also certainly broader than in the typical form. 
Although some Coleopterists have followed Thomson in regarding H. picicrus as 
specifically distinct, others, and probably the majority, have regarded it merely as a 
variety of H. fuscipes. 

After carefully comparing a number of specimens of H. fuscipes and picicrus, I am 
inclined to think the latter is not entitled to specific rank, chiefly because there are all 
grades of intermediate forms between the two extremes. H. picicrus seems to prefer 
a different environment from H. fuscipes, although there are perhaps few places where 
either form exists to the complete exclusion of the other. H. picicrus is more common 
on peat mosses than is H. fuscipes, while it is certainly much scarcer on salt-marsh 
areas than is the latter. I think also H. fuscipes is less common in the north and 
picicrus less common in the south, but of this T am not altogether certain, and even if 
it is so, the distribution may be owing to the distribution of peat mosses. 

Characters of the Imago. 

Without entering into a detailed description of the imago, I wish to refer to certain 
characters which are of special interest. The head bears the two antennae which are 
of the type found in all the Hydrophilidae (fig. 34). They consist of nine segments, 
the basal one long and slightly curved, the next four short and simple. The sixth from 
the base is somewhat saucer-shaped, the concave surface being outwards and the attach- 
ment being central. Above the saucer-shaped segment are two more or less gobular 
segments each flattened above, and upon them is the apical gobular segment. The 
three terminal segments are dark brown in colour and are thickly covered with short 
stout hairs, while the upper face of the saucer-like segment bears a number of short 
tubercles from the apex of each of which stands up a long fine hair. The colour of this 
segment, and of all those below it, is pale testaceous. 

The antennae are attached to the head immediately in front of each eye, and when 
at rest lie backwards along the ventral margin of each eye and just below the sides of 
the head. 



The Mouth Parts. 

The labrum is long and narrow, being slightly indented on the free border in the 
middle. It can be seen moving up and down when the beetle is feeding. The right 
and left jaws (figs. 1 and 2 ; 32 and 33) differ somewhat from one another. They are 



THE LIFE-HISTORY OF HYDROBIVS FUSCIPES, L. 319 

of the type characteristic of the group, serving the double purpose of tearing and 
grinding. Seen from above, both jaws show a two-pointed apex, the outer point of the 
left jaw being slightly thickened and having also a double point slightly marked. On 
the inner edge of the left jaw just below the apex is a thick tuft of long bristles, and 
immediately below this tuft is a second series of bristles of which those nearer the apex 
are longer than those nearer the base. The uppermost bristle of this second series has 
a thickened base. 

The inner edge of the right jaw shows only a very small tuft of a few bristles in 
place of the upper tuft on the left jaw, but the second series is similar in both jaws. 
Seen from above, the base of each jaw ends in a triangular projection of which the 
surface is smooth, as is also the margin of that projection in the right jaw, but in the 
left the margin is " feathered " by a series of short hairs. Seen from below, the jaws do 
not show much more than when seen from above. They are, however, three-sided, possess- 
ing not only a dorsal and ventral but also an inner face, broad at the base and narrow- 
ing to a point at the apex ; and the inner face of the one jaw is, of course, opposed to 
the inner face of the other. The basal part of this inner face of the jaw is the grinding 
region, the food being brought here by the action of the maxillae and the points of the jaws 
and then ground down between the bases of the two jaws before being finally swallowed. 
And this grinding apparatus is very efficient, as the two halves of it differ from one 
another in accordance with different functions. The grinding region of the left jaw is 
somewhat concave, and has a series of strong ridges running down it parallel to the 
length of the mandible. These ridges do not cover the whole of the concavity, but taper 
off towards the dorsal side of it and give place to what appear to be a number of minute 
tubercles. The grinding region is fringed all round with short hairs; and it is these 
which are partly visible in the dorsal view of this mandible. 

On the right jaw the corresponding region is somewhat convex, obviously to fit the 
concavity on the left jaw. There are no fringing hairs or deep grooves or ridges, but 
the whole surface is sculptured like a fine file. 

In a general way this description would probably apply to the grinding apparatus 
of all the Hydrophilidae, but there are probably slight variations in the different genera. 
For instance, in Hydrophilus piceus, L. , most of the concavity of the left mandible is 
file-like in sculpture, the strong ridges being confined to a small portion of the upper 
border of the concavity. 

The maxillae (fig. 3) are brush-like organs. The galea of each is crowned by five or 
six rows of stiff hairs incurved towards the mouth. These rows are placed one behind 
the other and no doubt make a very effective brush. The lacinia also bears a strong 
tuft of hairs, and there are also stiff hairs on the free edge of the stipes along the upper 
third of its length. The rest of the free edge of this part gives off a very fine trans- 
parent membrane as a narrow strip along it. The palpi are long and composed of four 
simple segments, the basal one being very short. 

The labium (fig. 4) has the galeae and lacineae feathered along their free edges with 



320 MR FRANK BALFOUR-BROWNE ON 

fairly long hairs forming a more or less continuous row right across the front. The 
palpi are three-segmented, the middle segment of each bearing a thin row of long hairs up 
its inner side, while the apical segment bears one or two very long hairs on its outer 
side about one-third of its length from the apex. 

When the beetle is out of the water the maxillary palpi are withdrawn and lie 
folded against the maxillae under the sides of the head, the antennae being turned 
forward and constantly waved about. In the water, on the other hand, the antennas 
are withdrawn and the maxillary palpi extended, obviously being the subaquatic organs 
of touch and possibly also of smell and hearing. 

The Thorax. 

Only two or three points about the thorax call for remark. The mesosternum bears 
in the median line a small pointed projection such as is found in certain other members 
of the Hydrophilidse, e.g. Anaccena limbata, F. ; but whereas it is in the genus Anacsena 
useful for distinguishing A. limbata, F., and ovata, Reiche, from A. globulus, Payk., 
where it is represented only by a small tubercle, it does not help in Hydrobius to 
distinguish H. picicrus from H fuscipes. 

The three pairs of legs have the femora covered with a dense mass of short hairs 
except in the apical region, which is smooth and shining. 

On the mesothorax and attached both to the tergite and to the elytra is a pair of 
small alulae. It has been suggested that these organs, which occur in the Dytisoidae 
and Hydrophilidse, may be responsible for the sounds which some of the species produce, 
but this suggestion is discounted by the fact that, although both sexes of Hydrobius 
fuscipes possess them, only the male produces a noise. 

With regard to the abdomen, the only part of special interest for this paper is the 
sexual armature, especially that of the female. 

The Male Sexual Armature. 

The penis (figs. 10 and 11) is of the general form common to all the Hydrophilidae, 
possessing two side pieces or valves which are said to be at least partly for the purpose 
of retaining the penis within the vagina at the time of copulation (Henneguey, Les 
Insecles, 1904, p. 185). I am inclined to think that they are rather for the purpose 
of spreading the walls of the vagina so as to allow the seminal fluid to escape rapidly 
from the duct, and no doubt the valves also function as a penis sheath when in a state of 
rest. I have used the term " penis " in connection with this organ, but it is perhaps open 
to question whether it is correctly so used. Some authors speak of it as " the sedeagus." 

There seems to be some difference of opinion as to what abdominal segments are 
represented in the male of Hydropldlus, and since Hydrobius is very similar to Hydro- 
philus as to its male sexual armature, the remarks apply equally to both genera. 



THE LIFE-HISTORY OF HTDROBIUS FUSCIPES, L. 321 

Packard (1898, p. 181), although not describing his views, gives the figures of Kolbe (vide 
figs. 203 and 204) according to which the sclerites of segment 8 have disappeared, the 
segment being only represented by a length of connective tissue between the sclerites 
of segments 7 and 9. Peytoureau (1895), on the other hand, only recognises as 
segments those regions bearing one or more sclerites, and he considers segment 8 to be 
represented by both tergite and sternite, while segment 9, the terminal one, is represented 
only by a sternite. Packard treats Peytoureau's segment 8 as being really the tergite 
of segment 9 (which he considers has no sternite) and the sternite of segment 10, which 
has no tergite, and he calls Peytoureau's 9th sternite the external lobes of the penis. 

Henneguey (1904, p. 185) has followed Peytoureau, as he states that the repro- 
ductive organs of the males of Coleoptera are always found upon the posterior border of 
the 9th sternite, and he gives one of Peytoureau's figures. It is difficult to find any 
justification for the view of Kolbe and Packard, since the connective tissue region 
which they treat as segment 8 is no better marked than a similar region between their 
segment 10 and the sclerites of the penis. There is, however, this to be said for their 
view, the sclerite they describe as the external lobes of the penis, which is Peytoureau's 
9th sternite, certainly functions as part of the penis. In copulation the posterior 
segments of the abdomen and also the penis — which in a state of rest are telescoped 
within the other abdominal segments — are protruded, and the penis, together with the 
external lobes (Kolbe), i.e. the 9th sternite (Peytoureau) is bent downwards and to 
the right or left according to the position of the male on the back of the female, and a 
distinct kink is visible at the bend. The region beyond the kink, i.e. penis and external 
lobes, bends right round until it comes to lie beneath the abdomen ; that is, it turns 
through an angle of 180° and the dorsal side remains dorsal during copulation. 
Mr Donisthorpe (Ent. Rec, xii. 291, 1900) observed the same fact in Hydrophilus 
piceus, L. 

The Female Sexual Armature. (Pigs. 12-14.) 

Peytoureau (I.e., plate xx.) gives a drawing of the female armature of Hydrophilus 
in what he describes as a state of activity, and he says (I.e., p. 154) that the 8th 
sternite " porte a son bord posterieur mamelonne deux apophyses simples, courtes, 
dirigees en arriere, armies a leurs extremites de poils sensitifs, cachees par le huitieme 
urite et deux autres apophyses laterales plus longues egalement terminees par des poils 
sensitifs " ; and he further says (I.e., p. 155), " C'est apres le septieme sternite que s'ouvre 
l'oviducte simplement membraneux. L'articulation tergo-sternale du huitieme urite a 
lieu au point x (plate xx. fig. 3) ; elle est assez lache, grace a la presence d'une 
membrane d'union." 

If this is a correct statement for Hydrophilus, then Hydrobius appears to differ 
very decidedly, as, from my observations, it is difficult not to associate the armature 
with the 8th rather than with the 7th segment — a condition which, according to 
Peytoureau, exists in Dytiseus. 



322 MR FRANK BALFOUR-BROWNE ON 

Segment 8 is represented by a partially chitinised tergite, immediately underneath 
which opens the anus. On either side of this tergite (8'), extending round the sides 
of the segment, is a narrow chitinous band (y). The two bands extend to the ventral 
face of the segment and then turn through about a right angle and pass backward, 
lying more or less parallel to one another. These chitinised bands do not seem to 
represent any regular sclerite of the segment, but are rather a specialised carrier 
for the sexual armature which is pivoted on the apices of the two parallel ventral 
chitinous rods (xx). 

The bands are, however, undoubtedly part of the 8th segment, and the part which, 
in Hydrophilus, Peytoureau regards as the 8th tergite (c), is situated posteriorly 
with regard to them. 

The sexual armature consists of two swollen masses (aa) from the upper side of each 
of which a long tapering filament (bb) about 1 mm. in length projects backwards. 
These two filaments, which correspond with Peytoureau's " deux autres apophyses 
laterales plus longues . . .", are the spinnerets, and consist each of three portions, a 
long basal segment and a short apical one bearing a long fine and tapering hair. On the 
ventral side of the armature, between the swollen bases of the spinnerets, projects a thin 
chitinised membrane (c) — (which, from Peytoureau's drawing (plate xx. fig. 3, viii.), I 
gather he regards as the sternite of the segment) — which, in a ventral view, has a deep 
groove running up its median line on either side of which the membrane fits round the 
base of a spinneret, the middle region of the posterior free margin projecting into 
two horn-like processes between the spinnerets and one on either side of the median 
groove. The posterior free edge of this membrane is fringed with hairs and the two 
horn-like processes are Peytoureau's " deux apophyses, simples, courtes dirige'es en 
arriere," etc. 

The vagina opens between the spinnerets and just dorsal to this ventral chitinous 
plate, so that if Peytoureau is correct in describing this ventral plate as the 8th sternite, 
the genital aperture is, on his own showing, not between the 7th and 8th segments, but 
posterior to the 8th. 

I have said that the armature is pivoted on the two chitinous rods (xx) previously 
described. When the genital armature is in a state of rest, the swollen bases of the 
spinnerets are drawn back by muscles attached to them dorsally, by reason of which 
they turn on the two pivots already referred to, and under these circumstances the 
spinnerets are concealed for about half their length by the grooved ventral plate which 
wraps round them, since the ventral plate is fixed and does not move on the pivots 
(vide fig. 14). When the armature is in a position of activity the bases of the spinnerets 
are thrown forward on the pivots and the spinnerets thus project farther beyond the 
grooved ventral plate. The pivots upon which the spinnerets swing are apparently 
what Peytoureau describes as " l'articulation tergo-sternale du huitieme urite. . . ." 
Peytoureau's figure (plate xx. fig. 1) represents what appears to me to be a dis- 
section of the parts in a state of rest. 



THE LIFE-HISTORY OF HYDROBIUS FVSCIPES, L. 323 

Secondary Sexual Characters. 

There seems to be no character by which the sexes can be distinguished at first 
glance, in fact the only differences I have been able to detect between the sexes are that 
the antennae and maxillary palpi of the male, when compared with those of the female, 
are distinctly stouter. In Hydrophilus the sexes are easily distinguished by the anterior 
tarsi, which are simple in the female and have a dilated apical segment in the male. 
In Hydrocharis the male has the claws of the anterior tarsi " sharply bent like a 
grappling hook," * but there is no such distinction in Hydrobius fuscipes. 

Habits. 

The food of the imago Hydrobivs consists chiefly of plant tissues, although it is also 
very ready to eat up dead insect larvae, snails, etc. Its favourite food appears to be the 
filamentous algae and also algae growing upon stems of other water-plants, etc. It also 
readily devours dead plant tissues, for instance, decaying blades of grass, leaving the 
fibrovascular bundles, or at any rate stripping them first of the softer tissues. Apparently 
only when driven to it for want of other food will it attack living higher plants, and 
then it seems to prefer the epidermis. It was quite extraordinary the amount of 
filamentous algae which it caused to disappear when kept in an aquarium. 

During the colder months of the year Hydrobius disappears more or less completely 
from its usual habitats but, as with regard to other water-beetles, there seems to be no 
definite knowledge as to where it disappears to. The beetle passes the winter as an imago. 
I have never found the larvae in the autumn, although Miall (1895, p. 93) says that 
eggs and larvae are to be found in August. I have, however, occasionally found the eggs 
of Helochares, a related genus, in August, but I fancy that autumn eggs are exceptional 
and that the larvae do not survive the winter. I think that the imago burrows in the 
mud at the bottom of the pond and passes the winter in the torpid state, but this is only 
theory. The only evidence I have is that in the aquaria in cold weather very few 
beetles are visible, and as they cannot get out of the water they must be amongst the 
stones at the bottom. Also some specimens of Bidessus minutissimus, which I have 
kept through the winter in a tumbler, have spent the whole time amongst the fine 
gravel, and I have not seen them once for many weeks. 

With regard to breeding habits the usual egg-laying period seems to be from March 
to July, although Miall (I.e., p. 93) mentions eggs in August. This year I found several 
egg-cocoons on 2 1st February, but I think that was exceptionally early owing to the long 
spell of mild weather in this district (near Belfast). Seeing that the larval period lasts 
from twelve to fourteen weeks, I doubt if eggs laid later than the end of July would 
have much chance of getting through before the cold weather, and the absence of larvae 
during the winter seems further evidence of the limits of the egg-laying period. 

* Fowler, W. W., British Goleoptera, i., p. 220. 



324 MR FRANK BALFOUR-BROWNE ON 

I am not sure as to when the males reach maturity, but they are sexually excited 
early in February. Possibly the females are " fertilised " in the autumn, as, so far as I 
have observed, they will not take a male at the beginning of the season. The only time 
I have seen them accept a male is immediately after having completed an egg- cocoon. 
Donisthorpe (I.e., p. 291.) refers to the refusal of the female Hydrophilus to accept a 
male as a case of sexual selection. So far as my experience goes, it is not a question of 
the choice of a male by the female, but of the male courting the female at the wrong 
time, and if a female has her spermotheca full of spermatozoa, she wi]l not unnaturally 
reject any male. Immediately after completing an egg-cocoon a Hydrobius will accept 
any male which offers, and I expect the same is the case with Hydrophilus. 

The courtship is a somewhat strange phenomenon, and it is the only period at which 
I have heard the male stridulate. I think that the males are only excited during the 
daytime, at least in the earlier months of the year, and require suitable light and 
warmth to arouse them. For instance, in February or early March in my aquaria, kept 
in a warm room, there are few beetles visible at 9 a.m., most of them being amongst 
the stones at the bottom. About 11 or 11.30 the sun reaches the aquaria, and then 
all the beetles come to the surface and move about rapidly and feed and the males are 
then excited. If at 9 a.m. the temperature of the water is raised from 50° F. to about 
60°, the same phenomena are observed as occur later in the morning under the 
influence of the sunshine. 

The male sits on the back of the female rather in the position of a jockey on a race- 
horse, that is, well forward, his metasternum being about over her prothorax. His 
head is bent right over hers, and his courtship consists in drumming upon her labium 
with his maxillary and labial palpi. His maxillary palpi pass round outside the bases 
of hers. She bends her head downward as if trying to avoid his attentions, and her 
maxillary pa]pi wave about. Her jaws can be seen working, and she occasionally 
succeeds in biting the palpi of the male. The front tarsal claws of the male hold on 
either in front of the eyes or behind them or even to the eyes themselves ; the mid- 
tarsal claws catch the posterior margin of the female prothorax, while the posterior tarsal 
claws hold the borders of the elytra. 

If the female is very annoyed by the attentions of the male she signifies it by 
swaying rapidly from side to side, or by running quickly over the submerged vegetation, 
and, as a rule, after a short struggle the male desists. Otherwise she may feed quietly, 
quite regardless of his wooing. He makes occasional attempts at copulation by rapidly 
moving backwards and extending his abdominal segments and the penis, but the 
female usually sweeps the latter away with her posterior tarsi. After making several 
ineffectual attempts the male usually desists and goes in search of another mate. 

During courtship the male keeps up a faint but distinct stridulation, reminding one 
of a corncrake at a great distance. At the moment the noise is produced the ventral 
side of the abdomen collapses like bellows, as if expressing air, and at each collapse of the 
abdomen the apex bends downwards. I can find no stridulating file. The noise is not 



THE LIFE-HISTORY OF HYDROBIUS FUSCIPES, L. 325 

produced by the rubbing of the tips of the elytra against the abdominal tergites, as 
these regions are quite smooth. The other British species of the genus, Hydrobius 
oblongus, Herbst, stridulates much more distinctly, and I believe in that case the 
noise is made by both sexes, and it certainly is not specially confined to the periods of 
courting. 

The Spinning of the Egg-Cocoon and Oviposition. 

The cocoon of Hydrobius differs somewhat in form from that of Hydrophilus, and 
Lyonnet's description of the method by which the female of the latter spins her cocoon 
would not apply to the female Hydrobius. Whereas Hydrophilus builds a more or less 
spherical cocoon with a short "mast" projecting unsupported from the water, Hydrobius 
builds a cocoon attached along one side to its support, and with a flat, ribbon-like process 
extending to the water surface in place of the " mast." 

According to Lyonnet (1829-32), Hydrophilus builds her cocoon in two parts ; that 
is, having built so much, she deliberately turns round and constructs a replica of the 
first part and then weaves the two parts together. Under normal circumstances 
Hydrobius builds her cocoon straight through without turning round. 

The favourite situation for the cocoon is on the under side of a growing blade of 
grass floating upon the water. The female first takes up a large supply of air so that 
her subelytral space is full, and she has also a large bubble attached to her under side. 
She lies lengthwise along the under side of the blade of grass, and, projecting her 
spinnerets fully, she brings them and the air-bubble into contact with the grass 
and the silk begins to flow as a fine thread from the apex of each of the spinnerets. 
The whole cocoon is formed on the film of the bubble which is on the under side of 
the insect. The spinnerets work the whole time in the air, pressing against the film 
and weaving the cocoon on its curved surface. 

The silk will not form threads in water, but comes away from or adheres to the 
apices of the spinnerets in the form of whitish lumps. It seems to be unusual for 
the beetle to build the cocoon far below the water. I tried the experiment of submerging 
the grass blade while the cocoon was being built. After a time the beetle turned to 
the side of the blade to reach the surface, but failing, she continued to work for a minute 
or so. She then made more determined attempts, and as she was in danger of asphyxiat- 
ing, she very reluctantly broke the air-bubble which united her with the cocoon and 
reached the surface. Having renewed her supply she returned to the cocoon, and by 
moving her abdomen over it endeavoured once more to take up her work, but after two 
or three failures she gave it up and later started a new cocoon nearer the surface. 

The silk threads are first spun backwards and forwards across the grass blade and 
the insect then begins to spin on the film immediately behind her and so raises a wall 
which soon takes the shape of a blunt cone.- She gradually moves forward, spinning all 
round the film, and thus extending the cone. She never turns over, but continues facing 
the same direction all the time, occasionally turning to the side of the grass blade and 
TRANS. ROY. SOC. EDIN., VOL. XLVII. PART II. (NO. 14). 50 



326 MR FRANK BALFOUR-BROWNE ON 

elevating her head to the surface and renewing her air-supply. She also occasionally 
slightly raises her elytra and separates them and then closes them down again, thus 
enlarging her ventral bubble from her subelytral air-supply. 

The building of the body of the cocoon occupies an hour or more, but oviposition 
always commences before the body is complete and often quite early in the building 
process. I removed one female after she had been spinning for twenty minutes, by which 
time she had only formed a small cone of silk and she had already deposited three eggs. 
An egg passes rapidly from the oviduct, scarcely interrupting the spinning. It passes 
out between the spinnerets and rubs against the fringing hairs of the grooved ventral 
plate. The spinnerets deliberately place it in the cocoon, and then rapidly spin threads 
backwards and forwards over it, fastening it down. The spinnerets are remarkably 
sensitive. On several occasions T ripped gashes in partially formed cocoons and the 
spinnerets quickly turned in the direction of the rent and closed it up. 

The number of eggs laid in a cocoon is from thirteen to nineteen, and possibly may 
vary beyond these figures. Once they are laid the female spins a layer of silk over 
the top and then continues the sides of the cocoon, leaving a certain amount of space 
unoccupied by eggs but filled with air. From time to time she squeezes the cocoon 
with her hind tibia?, shaping it ; and, as a rule, this squeezing is specially noticeable where 
the cocoon ends off with a more or less flattened surface rather like a lid in appearance. 

Once this flat surface is complete the female continues to spin a narrow band on the 
air-bubble, every now and again pressing it against the grass blade and thus attaching 
it. She carries this ribbon to the edge of the blade and then spreads it out on the water 
and on the upper surface of the grass. Its form and length depend upon the position of 
the cocoon with regard to the surface of the water, and sometimes, instead of being 
spread out on the surface, it is carried up the support into the air, but under normal 
circumstances the ribbon always reaches the surface. 

As to the process of spinning, the silk comes off from the apex of each spinneret as 
a fine thread, and the two spinnerets work side by side backwards and forwards, attach- 
ing the threads at the end of each stroke. But although the silk thread comes from 
the apex of the spinneret there is not, so far as I can discover, any canal running up 
within that organ. The long apical hair, although not solid, has no aperture from 
which the silk could flow, and the apical and basal segments are, I believe, filled up with 
tissues. 

So far as I can make out, the silk flows along the outer surface of the spinnerets, 
which appear wet all the time that spinning is going on. During spinning the 
spinnerets are fully extended and the grooved ventral plate fits around their bases. The 
fringe of hairs on the free edge of this plate appears, like the spinnerets, to be wet, and 
I think the silk fluid flows off the two horn-like processes of the plate which are pressed 
between the spinnerets and runs out along these. 

I found that once the beetle had fairly started making the cocoon it was possible to 
lift the grass blade completely out of the water and even to place it under the micro- 



THE LIFE-HISTORY OF HYDROBIUS FUSCIPES, L. 327 

scope without disturbing her. The cocoon is formed just as easily under these circum- 
stances as under water, provided that the temperature is not too high or as long as there 
is sufficient moisture in the surrounding air. I found by placing a spinning female in 
the direct sunshine that, after a short time, the spinnerets, although continuing to work 
backwards and forwards, produced no silk. At first the silk spun became thicker than 
usual and then it became lumpy and stuck to the spinnerets, and I think this was due 
to the heat drying up the silk fluid on the spinnerets. It must be admitted that if this 
explanation is correct it is difficult to explain how, when the beetle is upside down and 
working on the under side of the grass, the silk runs out the length of the spinnerets 
against gravity ! 

About half way up the basal segment of each spinneret and on its inner face is a 
transparent region of more or less oval shape. This region is beyond the "horns" of 
the grooved ventral plate when the spinnerets are extended, but is completely covered 
by them when these organs are withdrawn. Under high magnification ( x 400) this 
region shows a structure different from that of the rest of the spinneret, and I at first 
thought that the ducts of the silk glands might open here, but I cannot detect any 
pores. I think therefore that the silk glands must open to the exterior somewhere 
below the fringe of the ventral plate, that the silk flows out on the hairs of this and 
passes on to the spinnerets from them. 

The posterior end of the abdomen of the female is extremely flexible during the 
spinning of the cocoon, segment 7 moving very freely upon segment 6, and the 
spinnerets also waving about with independent movement. 

As a rule, shortly before a female has completed her cocoon, a male takes up his 
position on her back and commences stroking her labium. Within two minutes of 
completing her cocoon the female accepts the male and copulation lasts about a minute. 
The male ceases to stridulate the moment he begins to protrude the penis. 

The Development of the Egg. (Figs. 15-31.) 

The development of the egg of Hydrophilus has been described by several authorities 
such as Heider, Graber, Kowalevsky, etc., and that of Hydrobius appears to be very 
similar, but Hydrophilus has been studied chiefly from the histological point of view, 
and I propose to describe the development of the embyro of Hydrobius as seen from 
day to day through the shell of the living egg. There are certain points which are not 
referred to by the above authors which are of some interest. 

The egg of Hydrobius is oval in shape, from l'l to 1*4 mm. in length, and bears at 
one end a small pedicel, this being the end of the egg at which the head of the embryo 
develops. The newly laid living egg shows no structure under the microscope beyond a 
shell membrane and granular contents which completely fill the shell. After twenty-four 
hours the contents have shrunk so that there is now a broad, perivitelline space, and on 
the contents there are at this time certain irregularities which suggest that cell-division 



328 MR FRANK BALFOUR-BROWNE ON 

has commenced. By careful focussing and manipulation of the light a median groove, 
the " primitive streak," is to be seen, which extends for about half the length of the 
egg from the posterior end. 

On the third day the groove is more visible, and at the posterior end of the egg the 
commencement of the amnion and the amniotic cavity is apparent. There are also 
indistinct traces of segmentation of the yolk, and twenty-four hours later these have 
developed into complete segmentation. By this time — the fourth day — the amniotic 
fold has extended forwards over the median groove, but has not yet completely enclosed 
it, so that the blastopore is to be made out over certain irregular outgrowths which 
constitute the beginnings of the embryonic head. The groove has extended round the 
posterior end of the yolk and along the dorsal side so that the " tail" end of the 
developing embryo lies almost over the head. Indistinct traces of somites are 
visible in the embryo which, by transmitted light, is more transparent and less 
granular than the yolk. 

On the fifth day the form of the embryo is quite distinct, the yolk having consider- 
ably diminished. The embryo lies round the yolk, its ventral side towards the shell. 
The appendages of the head, thorax, and abdomen are recognisable as buds. It is to be 
noticed that the tail end of the embryo does not now extend quite so far round the yolk 
as it did in the previous stage. 

On the sixth day the mouth parts, the maxillae, and the labium (which at present 
consists of two separate parts) have folded on themselves and the maxillary palpi are 
recognisable as buds. The embryo has still further contracted in length. The legs, 
although recognisable, are at present only partly defined, but on the seventh day they 
are more distinct. The segmentation of the abdomen, which on the sixth day has only 
just commenced, is also more distinct on the seventh day, and the embryo is still shorter 
in length. 

In the fifth-, sixth-, and seventh- day stages the embryo is seen to be surrounded by a 
membrane— the amnion — within the shell. On the eighth day this has disappeared — it 
having burst along the ventral side and shrunk back to lie as an irregular double ridge 
on the dorsum. It goes by the name of "the dorsal organ" or "dorsal tube." The 
embryo has now contracted so much that it lies straight out within the shell. At this 
stage the eyes first become visible as six small pigment spots on each side of the head. 
The labium is now a lilobed structure lying backwards on the ventral side of the head, 
and the maxillae lie in the same direction on either side of it. 

During the next twenty-four hours the abdomen begins to lengthen out, but this 
time it curves ventrally instead of dorsally. The legs, which up to now have lain close 
together in the median line, have separated, leaving a space between those of the two 
sides of the body. The "dorsal organ" has by the ninth day completely disappeared, 
having sunk into the yolk and been absorbed. As soon as the dorsal organ disappears 
the dorsum of the embryo closes completely over the yolk, which up to that time has 
remained exposed. The trophi have lengthened and show traces of segmentation. 



THE LIFE-HISTORY OF HYDROBIUS FUSGIPES, L. 329 

Within the posterior abdominal segments are traces of the lateral tracheal trunks 
which are just forming. By this time all trace of the abdominal appendages has 
disappeared. 

On the tenth day the trophi and legs show segmentation and have lengthened out 
considerably, while the segmentation of the body somites is also well advanced, proceed- 
ing forwards from the posterior end of the abdomen. The lateral tracheal trunks are 
now quite distinct in the posterior part of the abdomen as colourless tubes. 

The eleventh day seems to be the muscle-forming period, as at this stage muscle 
fibres are visible in various parts of the body. The lateral tracheae which have not yet 
developed through the whole length of the embryo show, posteriorly, rudimentary 
branches, and in the median dorsal line faint traces of the heart are visible in the 
posterior segments. 

The most marked change on the twelth day is the extension forward and the 
pigmentation of the lateral tracheal trunks. These now reach into the head and show 
indications of the main branches and of the future " setose gills." 

On the thirteenth day the tracheal system is almost completed, the dorsal network 
having been formed, and the branches of the head can be seen tapering off to fine threads, 
but, so far as I can make out, the "setose gills"* themselves do not appear until the 
fourteenth day. There are, however, only small differences noticeable in the living egg 
of the thirteenth and the fourteenth day. 

The fourteenth day is the last stage, the embryo now being complete within the 
shell. One point to notice is that, almost from the commencement of development the 
embryo has never completely occupied the shell, there having all through been a fair 
amount of free space. On the fourteenth day, however, the embryo commences to swell 
up two or three hours previous to its escape, so that at the time of the bursting of the 
shell there is no visible unoccupied space within the shell. 

While watching this process of swelling up of the embryo I found in the head, 
situated in the median line and immediately below the dorsal aorta, a pulsating 
organ. It pulsates much more slowly than the heart, only commences to pulsate a few 
hours before the embryo escapes, and does not, so far as I have observed, pulsate con- 
tinuously from the time it first starts. It begins with a few beats, after which there is 
a pause of, may be, many minutes, but as the time of emergence approaches the pauses 
are fewer and the pulsations rather more rapid, being about 40 per minute, while the 
the heart is beating at quite 100 per minute. 

This pulsating organ is driving fluid backwards, but it was not possible to see where 
the fluid came from or was driven to. I discovered a ligament suspending the organ, 
and as it passes on one side of the aorta, I expect that it is one of a pair. 

There are no blood corpuscles visible in the fluid pumped through this pulsating 
organ. In conjunction with its pulsations certain movements of the gut are visible, and 
these become more noticeable as the beats of the pulsating organ become more frequent. 

* Vide infra. 



330 MR FRANK BALFOUR-BROWNE ON 

The position of this organ in the head (immediately below the aorta and just at 
the point where the aorta divides into two vessels passing down the frontal region), and the 
apparent relationship between its pulsations and the peristaltic contractions of the gut, 
lead me to think that there is no special embryonic organ, but that the pulsations are 
due to the working of a pharyngeal pump. The larva, as we shall see, feeds by suction, 
and there is no reason why the pharyngeal pump should not be of service before the 
embryo escapes from the egg. I imagine that the fluid which surrounds the embryo 
within the shell is taken in by the mouth and pumped by the pharynx into the gut. 
Thus the embryo swells up and space is afforded it within the shell by the removal of 
the surrounding fluid.* 

For a short time before the shell bursts the embryo is so tightly packed that it is 
difficult to make out the limits of head and tail (vide fig. 31). The head lies well into 
the anterior end of the egg, while the first, second, and third abdominal segments occupy 
the posterior end. The shape of the egg is often somewhat altered by the packing of the 
embryo which, although usually as described, is sometimes turned in other positions. 

The shell ultimately bursts somewhere in the region of the maxillary palpi and the 
larva immediately straightens out, the shell falling off its back as it does so. All the 
trophi and legs at once assume their normal positions, the labium folding forwards and 
coming to lie immediately behind the mouth, the maxillae likewise folding forward and 
lvino- on either side of the labium. 

I have described the maturation of the egg as taking fourteen days from the time 
it is laid until the escape of the larva. This seems to be the usual time in April or 
May, but the rate of development varies with the temperature. Thus eggs laid in 
February — which is exceptionally early — took twenty days to mature, while in one or 
two cases during the summer thirteen days were sufficient. 

Kowalevsky (187l) and Heider (1889) found that the eggs of Hydrophilus 
developed in twelve days, while Miger (1809) gives twelve to fourteen days as the usual 
period. In my experience with this species twenty days and more was the usual time 
from the building of the egg-cocoon to the appearance of the larva, but as my specimens 
were kept in the north of Ireland — far beyond the normal range of the species — this 
may account for the slower rate of development. 

I have not followed the development of the embryo of Hydrophilus, but the 
statements of Kowalevsky, etc., indicate that the stages are practically identical with 
those of Hydrobius. None of these authors, however, describe the swelling of the 
embryo prior to the bursting of the shell nor do they refer to the cephalic pulsating 
organ. 

The Larva. (Figs. 5 to 9.) 

The larva which emerges from the egg is a small white worm-like creature with a 
pale testaceous head and darker eyes, the eyes being the most visible part, and is about 

* On ihis point see " The Lii'e-history of the Agrionid Dragonfly," Proc. Zool. Soc. London, Aug. 1909. 



THE LIFE-HISTORY OF HYDROBIUS FUSC1PES, L. 331 

one and a half times the length of the egg from which it comes. But such larvae are 
never seen under normal circumstances. As it comes from the cocoon it is distinctly 
larger and darker and the head, with its extended jaws, seems also larger in proportion 
than in the newly hatched individual. Lyonnet states that the larvae of Hydrophilus 
remain a day in the cocoon before breaking out. Those of Hydrobius also remain in the 
cocoon for some time, but I do not know how long. They ultimately make their way 
out by breaking through the cocoon, usually along the edge of the flat, lid-like surface 
which is at the base of the ribbon and just about on the water-level. 

I hatched a number of larvae from eggs which had been taken from the cocoon and 
allowed to sink in a tumbler of water. The larvae crawled about aimlessly for two or 
three days, but a number of them died. I took some out of the tumbler and placed 
them in a drop of water upon a slide to examine them under the microscope, and they 
immediately began to crawl about and at last succeeded in getting their heads out of 
the water. Once this was accomplished the head and anterior end of the body was 
raised up, and the larvae remained in this attitude for thirty seconds or more. During this 
short time they visibly expanded, and I found, when I replaced them in the water, that 
they now floated at the surface instead of sinking. Examination under the microscope 
revealed that the gut contained through most of its length a series of air-bubbles. 
These larvae had, in fact, taken a "drink" of air at the first opportunity and had 
expanded in consequence, and I found by an examination of larvae of all stages that 
the gut almost always contained these air-bubbles. If a larva lying quietly out of the 
water or at the surface is suddenly submerged, it quite probably sinks, but if it is first 
gently disturbed it at once raises its head and takes an " air-drink," after which it is 
secure against drowning if it is placed in the water. 

From the fact that larvae can be taken unawares and sunk, it is obvious that the 
air in the gut disappears and has to be renewed from time to time, but whether such air 
is of any use in respiration I do not know. 

It is possible, then, to account for several things by this peculiarity of the larva. 
First, the air-space within the cocoon is probably for the purpose of enabling the larvae 
to get an " air-drink." A similar space occurs in the cocoon of Hydrophilus, and it has 
been suggested that it is to enable the cocoon to float. It no doubt makes the cocoon 
buoyant, but neither in the case of Hydrophilus nor in that of Hydrobius is the cocoon 
ever built free in the water ; it is always attached to some floating object. There 
seems, therefore, no good reason for regarding the air as specially for the purpose of 
buoying up the cocoon. It has also been suggested that the air is for the eggs, but the 
eggs of Hydrobius— and probably also those of Hydrophilus — hatch perfectly well if 
taken from the cocoon and submerged. 

Lyonnet observed that the larvae of Hydrophilus swell up three or four times 
the size of the egg from which they emerge before they have taken any food. This 
also is probably to be accounted for by assuming that they, like the larvae of Hydrobius, 
take an " air-drink " in the cocoon. 



332 MR FRANK BALFOUR-BROWNE ON 

We may, perhaps, also account for the "mast" on the cocoon of Hydrophilus and 
the ribbon on that of Hydrohius as connected with this peculiar habit of drinking air, as 
so far no satisfactory explanation has been given of these structures. They have been 
described as the result of excessive zeal on the part of the female — a sort of using up 
of silk, and the mast of the Hydrophilus cocoon has also been described as an air-funnel 
to aerate the eggs. The mast always projects above the water, while the ribbon is either 
spread out upon the surface or upon some object above the surface. 

The larvae of one batch of eggs do not all necessarily hatch at the same time, some 
of the eggs — probably the upper ones — hatching sooner than the others. Now the 
earliest hatched larvae presumably take their drink of air, and as they remain some time 
in the cocoon, they no doubt tend to use up the limited supply within the cocoon. 
When they escape they open the cocoon below or on the water-line, and there is there- 
fore a risk of the remaining air escaping. I think the mast and the ribbon are 
provided against all these possibilities, and to make certain of other larvae getting to 
the surface to obtain their supply of air. 

The head of the larva has two eyes made up of about six more or less oval patches 
of pigment. The antennas are short, three-segmented organs, the apical segment being 
much finer than the others. The other two segments are, in the youngest stage, about 
equal in length and twice that of the apical one, but in the growth of the antennae the 
basal segment grows more rapidly than the middle one, so that, in the final stage, the 
basal segment is about twice the length of the middle one. On the middle segment and 
just external to the insertion of the apical one is a small tubercle. The apex of the 
terminal segment bears two long hairs. 

The mandibles are large, and end in a sharp incurved point and bear three sharp 
teeth on their inner margin, the basal one being small. The two mandibles are quite 
similar to one another. The maxillae are long and are presumably to be described as 
consisting each of a cardo and stipes bearing a minute inner lobe — the undifferentiated 
galea and lacinia — and a large three-segmented palpus. The middle segment of the 
latter bears at its apex, just external to the insertion of the apical segment, a long 
fine hair. 

The labium consists of a broad submentum and mentum, the latter bearing a pair of 
two-segmented palpi, and between them a small process presumably corresponding to a 
ligula. The submentum on its upper surface bears series of closely set minute outwardly 
projecting points. 

The labrum is peculiar. Its free edge is strongly serrated in the middle, but the 
organ is not bilaterally symmetrical. The central part of the anterior margin bears 
about five teeth. Some slight variations occur in the arrangement of the teeth, and 
they also change slightly at the moults, but the general arrangement is that there are 
five teeth of which the four on the right are more or less similar and evenly spaced, 
while the one on the left is widely separated from the rest of the series. This want of 
bilateral symmetry of the labium is not unique in Hydrohius, as it occurs also in 



THE LIFE-HISTORY OF HYDROBIUS FUSCIPES, L. 333 

Helochares, where, however, it takes a slightly different form, and is perhaps to be found 
elsewhere. There seems to be no reason for such irregularity, which is quite constant, 
and it is difficult to see how it can be of any advantage to the larva. 

On the posterior region of the head, on that part which telescopes under the tergite 
of the first thoracic segment, are two small sclerites. 

The prothoracic segment is entirely chitinised, and has no characters calling for 
remark. The mesothoracic segment is much shorter, and has a large chitinised tergite. 
On either side of it and at its anterior border a small projection exists into which runs 
a tracheal tube. Packard (I.e., p. 471) mentions that the American species of Hydro- 
bius possess seven pairs of short " setose gills." The larva of H. fuscipes has, on either 
side of the mesothorax, this setose structure, and on either side of each of the first seven 
abdominal segments is a similar structure also containing a tracheal tube communicat- 
ing with the main lateral trunk. Hydrophilus also possesses the abdominal " setae," but 
1 am- not sure as to whether the mesothoracic pair are present. The larvae of both 
Hydrophilus and Hydrobius drown comparatively easily ; they are never long away 
from the surface of the water, and are constantly to be seen renewing their air-supply 
by the posterior spiracles, and I do not see why these structures, which may be aborted 
or closed stigmata, should be described as "gills." 

The dorsum of the metathoracic segment is less strongly chitinised than that of the 
mesothorax, and on either side of the median line are two minute papillae each giving off 
a long fine stiff hair. Both meso- and meta-thorax tend to be bilobed at the sides, the 
one lobe lying above the other. 

The first abdominal segment is as short as the meso- or meta-thorax, but is distin- 
guished from these — apart from the absence of legs — by possessing two well-marked, 
though small, dorsal sclerites. The tracheal papillae or setose structures already referred 
to project on either side of the dorsum and near the anterior margin, and below this the 
sides of the segment are distinctly bilobed, the lobes being one above the other and 
somewhat resembling parapodia in appearance. Immediately behind the two sclerites 
and nearer the posterior margin of the segment are two papillae each with a long hair, 
and a similar pair lie, one on either side, just behind the tracheal papillae. The other 
seven abdominal segments are all very much longer from in front backwards, and are 
apparently composed of at least two annuli. The first annulus of each segment is 
plain, possesses no sclerites and no papillae. There is a slight constriction at its junction 
with the posterior part of the segment. The posterior part of each segment from 2 
to 7 is exactly similar to the anterior abdominal segment, since it possesses two 
distinct sclerites, smaller, however, than those of the first abdominal segment, a pair of 
tracheal papillae, and two pairs of papillae bearing long hairs. The anterior annulus of 
each of these segments is apparently the one which is specially contractile. Each of 
the papillae is apparently slightly chitinised, and perhaps their bases are correctly 
described as additional sclerites. 

The second part of segment 8 is somewhat peculiar. It extends posteriorly into 
TRANS. ROY. SOC. EDIN., VOL. XLVII. PART£lI.;(NO. 14). 51 



334 MR FRANK BALFOUR-BROWNE ON 

a three-lobed expansion, below the middle lobe of which the anus opens, and above 
which is a depression bounded anteriorly by a more or less perpendicular wall of tissue 
in the form of a slightly lobed ridge. Right and left of this ridge and just posterior 
to it is a small lobe bearing a flattened chitinised projection on which are three long 
hairs. The trilobed expansion is fringed with several long hairs and bears on its upper 
surface, just at the junction of the middle with the lateral lobes, two dark-coloured 
papillae each with two long hairs at its apex and one long hair projecting from a small 
papilla on its outer side. # 

On the posterior face of the anterior ridge and at either side of the median line is a 
narrow aperture — the spiracle — leading into a small chitinous vestibule into which opens 
the lateral tracheal trunk. The trunk is constricted where it opens into the vestibule, 
and there appears to be some apparatus of small chitinous pieces controlling this opening. 

The lateral lobes and the trilobed part of the segment are capable of spreading 
outward and of closing in, thus enlarging and diminishing the dorsal depression accord- 
ing as the larva is obtaining air at the surface and holding on to the surface film or 
moving under water. 

On the dorsum of this second part of the last abdominal segment is a large semi- 
circular chitinised tergite, developed perhaps for the protection of the functional spiracles 
which open just below its posterior margin. 

The surface of the skin of the larva is covered with scattered hairs arising out of 
small papillae amongst which are numerous small pores leading into short pits each of 
which gives off into the skin numerous minute hair-like processes. Amongst these pits 
are still more numerous minute pointed pits which look very like minute inverted spines. 

The larvae are metapneustic, the only functional spiracles being those opening into 
the dorsal depression of the eighth abdominal segment. The expansion of the lobes of 
this segment enable the larva to hold on to the surface film, the rest of its body being 
submerged. By moving the body rapidly from side to side when thus attached to the 
surface, the larva moves fairly rapidly ; otherwise its only method of progress is by 
creeping on the water plants, as it has no power of free swimming like the young larva of 
Hydrophilus. 

The position of the larva in relation to the surface film has nothing to do with the 
state of buoyancy, since an individual with no air in its gut can and does travel suspended 
to the surface film as easily as one filled with air ; but whereas the former sinks if, from 
any cause, it loses its hold, the latter floats and can regain its attachment to the film. 

The method of feeding of the larva is peculiar. Its habit is to lie half out of the 

* Between these two papillae is a small median pit from which bubbles of air seem to come off if, for instance, the 
larva is placed in alcohol. I have seen the same phenomenon in the case of the larva of Hydrocharis caraboides, L., and 
it would seem to indicate that there is a median spiracle or pair of spiracles opening at this point, although I cannot 
see any trachea; leading there. But by putting the larva in water under the microscope and flooding with alcohol, air- 
bubblesseem to come oil" from numerous rdaces. For instance, in one or two cases I distinctly saw bubbles escaping 
from the tracheal papillae of the first abdominal segment, and in many cases bubbles come from the sides of the 
other segments, but apparently not from the tracheal papillae. Where these latter bubbles come from I cannot 
determine, and at present I can merely mention these observations, leaving the matter for a future occasion. 



THE LIFE-HISTORY OF HYDROBIUS FUSCIPES, L. 335 

water, head downwards, on the side of the tumbler, and with the jaws widely expanded. 
Under natural conditions it may be found in a similar position on projecting water 
plants, or it is often to be seen enveloped in spirogyra with only its head projecting out. 
Its food consists largely of Entomostraca, daphnids being the chief item, though it also 
attacks small insect larvae and is quite ready to devour one of its own kind. 

The prey sooner or later comes within reach of the jaws and is at once seized. The 
larva then rapidly wriggles backwards out of the water, and having once got to a fairly 
dry spot, it raises its head and thoracic segments off the substratum and, holding the 
prey with antennae, maxillae, and labium, it proceeds to puncture it through and through 
by opening and closing the jaws, turning the captive round and round by movements 
of the antennae and maxillae. The outwardly projecting points on the upper surface of 
the submentum and the row of small spines on. the cardo of the maxilla assist in holding 
the prey and prevent the solid parts from being sucked down, as only the juices, with a 
large quantity of air, are absorbed. When there is nothing left but a crumpled mass of 
solid substance the larva lays this down beside it and either rests for a while or returns 
to the water in search of more food. It is apparently absolutely essential that the 
larva should get at least its head out of the water for feeding. By preventing larvae 
from so doing I found that they died in spite of being in the midst of daphnids. They 
would seize one and carry it about for quite a long time, always endeavouring to get out 
of the water. 

Quite recently I discovered a larva with its head buried in an egg-cocoon devouring 
a batch of fresh-laid eggs, but whether this is a usual occurrence or was a special 
discovery on the part of this individual I do not know. The cocoon was floating on 
the water, but I imagine that the eggs must have been submerged by the time the larva 
had penetrated the wall of the cocoon. 

The larvae seem to spend a great deal of their time out of water. I have often found 
them on the muddy sides of ponds, where they must be largely devoured by small birds 
and in the tumblers they were constantly to be seen on the sides. I have seen them 
seizing one of their own kind when out of water, but as a rule they return to the water 
for their food, and if they wander too far from it they are apt to suffer from drought. 

The duration of the first larval stage varies considerably. In specimens hatched in 
the middle of April the usual time until the first moult was about fifteen days, but later 
on, in warmer weather, the first stage only lasted from six to ten days. A second 
moult takes place at the end of a period about as long as the first. The two moults 
have, as far as external characters are concerned, produced no change in the appearance 
of the larva. In the young larva the head and jaws are much larger in proportion to 
the body than they are in the older larva, but this change is merely due to a more 
rapid growth of the body as compared with that of the head and is not connected 
with the moults. 

Having moulted twice the larva is in its last stage, and this occupies five to seven 
or even eight weeks. At the end of that time the full-grown larva crawls out of the 



336 MR FRANK BALFOUR-BROWNE ON 

water and burrows into the ground and after several days becomes a pupa. My larvae, 
reared in tumblers, indicated their maturity by great activity and by many futile 
attempts to get out. As soon as they were placed in small glass dishes containing 
moist earth they burrowed into it. 

In the earth the larvae spins a silken cocoon. My larvae burrowed to the bottom of 
the glass dishes in perhaps an inch of soil, and there formed the cocoons in contact with 
the glass. In some cases these were sufficiently thin to allow the larva to be seen 
within. I was surprised at the little difficulty I had in getting the larvae to burrow 
into the soil. Lyonnet found the larvae of Hydrophilus very particular in the choice of 
soil for pupating in, and only succeeded in getting pupa after several attempts. So 
far I have always failed to get the larvae of Hydrophilus to pupate, although possibly 
the fact that I have reared them beyond the limits of natural distribution may have 
affected them. The larvae of Hydrobius accepted garden soil without demur. 

Once the cocoon is complete the larva rests and does not at once change to a pupa, 
as I have found larvae unchanged after eight or nine days in the cell. 

The Pupa. 

The pupa is of the typical Coleopterous form, but bears on the dorsal side of its 
thoracic and abdominal segments series of stiff hairs. In Hydrophilus only a few hooked 
spines exist in place of these hairs, and Lyonnet discovered that they were supports 
upon which the pupa rests within the cell. The Hydrobius pupa, I found, was very 
active and turned in all positions in the cells, not by any means always lying passively 
on its back. 

The pupal stage lasts but a short time. I cannot give the exact duration, as I 
omitted to note the times at which the larvae changed to pupae, but, from the time 
the larva burrowed into the ground, it was three weeks before I saw any imagines in the 
cells, and as at least some of the larvae remained unchanged for eight or nine days, the 
pupal condition lasted about eleven or twelve days. There are, no doubt, variations in 
the length of this stage, just as there are in that of the maturation of the egg and of 
the larva, according to temperature conditions. 

The Life- Cycle. 

Miall (I.e., p. 93) states that "very soon after the emergence of the beetle the 
cocoon is formed and the eggs are laid," and " all the stages may be found together 
about the end of August." So far as my experience goes, Hydrobius completes its life- 
cycle during the one season, the whole cycle occupying about twelve weeks. I have 
never been able to find eggs or young larvae in August or later, yet, if Miall's state- 
ment is correct, larvae should be found during the winter, as are the larvae of the 
Dytiscidyj. It is possible that some of the later larvae may pupate in September and 



THE LIFE-HISTORY OF HYDROBIUS FUSCIPES, L. 337 

that the pupae may lie dormant through the winter, or possibly the imago may emerge 
from the pupal skin in the autumn and remain within the cell until the spring. Under 
ordinary circumstances the beetle remains some time within the cell before breaking out, 
but a newly emerged beetle is always recognisable on account of its comparative soft- 
ness, and I have no record of having found newly hatched imagines before June. 

Whereas a large number, perhaps all of the Dytiscida3, pass the winter in the larval 
as well as in the imago condition, I think that normally the Hydrophilidse are then only 
imagines, and that they all usually complete their life-cycle within a single season. 
This certainly is generally the case with Hydrophilus and, I think, with Philhydrus 
maritimus, Thorns, and, were it not for Miall's statement, I should also include Hydro- 
bius fuscipes, L. 

Fecundity. 

The egg-laying period of a single female apparently continues through the egg- 
laying period of the species. I have isolated a number of females from the time they 
built their first cocoon, and up to the end of June some have built as many as seven 
egg-cocoons. In these cases no male has been with the females since they started their 
first cocoon, with one or two exceptions, in which they received a $ after the completion 
of the first cocoon. The average number of eggs in these later cocoons is rather smaller 
than in the earlier ones, there being only about twelve ; but each of these females has 
produced from eighty to a hundred eggs up to the end of June, and there may yet be 
one or two cocoons built by them before the end of the season. The eggs are fertile, 
the development of the embryos being quite normal. 

Hydrobius fuscipes is certainly a common species, but there must be a tremendous 
mortality among the larva?. 

Summary. 

Hydrobius fuscipes, L., is a common water-bettle of which H. picicrus, Thorns, is 
apparently only a variety, the two forms being extremes of a series and tending to prefer 
different environments. 

The food of the imago is chiefly filamentous and confervoid algse, but it also 
devours dead insects, etc. The jaws, as in all the Hydrophilidse, have grinding 
surfaces at their bases, but the two jaws differ from one another in accordance with slight 
differences in function. 

The courtship of Hydrobius is peculiar, the male on the back of the female bending 
forward over her head and tapping upon her labium with his palpi. There are 
apparently only certain times at which a female will accept a male. The peculiar 
twisting of the penis in the process of copulation, which was observed in Hydrophilus 
by Donisthorpe, is also seen in Hydrobius. 



338 MR FRANK BALFOUR-BROWNE ON 

The female genital armature is apparently part of the eighth abdominal segment, 
so that Peytoureau's statement that, except in Dytiscus, in the Coleoptera this is always 
on the posterior margin of the seventh segment, is apparently incorrect. 

The spinning apparatus, which is somewhat different from that described for 
Hydrophilus, is described in some detail. The spinnerets, from the apices of which the 
silk runs, do not appear to have any duct running through them, and it is suggested that 
the silk flows out on their surfaces. 

The method of building the egg-cocoon seems to be quite different in Hydrobius from 
what it is in Hydrophilus, and whereas in the latter case it is stated (Lyonnet) that the 
eggs are laid after the body of the cocoon is completed, in Hydrobius, the egg-laying 
commences very soon after the cocoon is begun. 

The development of the embryo is followed in the living egg. In the final stage the 
embryo swells up so as to completely occupy the shell, and a cephalic pulsating organ, 
which is probably a pharyngeal pump, appears. It is suggested that this absorbs the 
periembryonal fluid and thus enables the embryo to completely occupy the shell. 

The larva shows certain peculiar habits, one of which is that it normally has its gut 
partly filled with air-bubbles, and as it takes its first " air-drink" shortly after hatching, it 
is suggested that the ribbon attached to the egg-cocoon, and which normally reaches 
the surface of the water, is for a pathway by which the larvse may get to the air. In 
the same way perhaps the mast of the Hydrophilus cocoon may be useful. 

Another peculiarity of the larva is that it can apparently only feed out of the water. 
The prey, which consists chiefly of daphnids, is sucked' dry, the solid portions being 
left as useless. 



BIBLIOGRAPHY. 



Heider, K., 1889, Die Embryonalentwicklung von HydropMlus piceus, L. I. vol., Jena. 

Henneguey, L. F., 1904, Les Insectes, Paris. 

Kowalevsky, A., 1871, " Embryologische Studien an Wurmen u. Arthropoden," Mem. Acad. St. Petersboury, 

(7), t. xvi. 
Laker, A. G., 1881, "The Cocoons of Hydrophilus piceus and Hydrobius fuscipes," Entomologist, xiv., 

pp. 82-84. 
Lyonnet, P., 1829-32, " Recherches sur l'anatomie et les metamorphoses de differentes especes dTnsectes," 

Mem. du Museum d'llist. Nat. Paris, xviii.-xx. 
Mi all, L. C, 1895, Natural History of Aquatic Insects, Macmillan &, Co. 
Miger, F., 1809, "Mem. sur les larves des Insectes Coleopteres aquatiques. I. Mem. sur le grand 

llydrophile," Assoc, du Museum d'llist. Nat., xiv. 
Packard, A. 8., 1898, Text-book of Entomology, Macmillan & Co. 
Pkytoureau, A., 1895, Morph. de VArmure genitale des Insectes, Paris. 
ScmdDTB, J. C, 1861-3, "De metamophosi Eleutheratormn Observationes. Bidrag til Inseckternes 

Udviklings-histoiie," Naturliistorisk Tidsskrift, Kjobenhavn. 



THE LIFE-HISTORY OF HYDROBIUS FUSCIPES, L. 339 

DESCRIPTION OF PLATES. 

Plate I. 

Fig. 1. Right mandible of imago, showing the grinding region (a). Partial view of the lower side 
and of the " inner " face. 

Fig. 2. Left mandible, showing concavity (b) into which the grinding portion of the right mandible 
fits. Partial view of the lower side and of the " inner " face. 

Fig. 3. First maxilla of left side of imago showing the series of "brushes" which assist in the feeding 
process, (a) cardo, (b) stipes, (c) palpiger, (d) galea, (e) lacinia, (/) the transparent membrane of the stipes. 
m., muscle attachments. 

Fig. 4. The labium from below, m., mentum ; sm., submentum. 

Fig. 5. Dorsal view of a larva recently hatched ( x 42) ; dotted portions are chitinised. 

Fig. 6 represents the maxillae and labium of larva as seen in a moulted skin by removing the frontal 
sclerite with the antennas and mandibles. Note the upper face of the submentum (s.) with its outwardly 
pointing teeth, m., first maxilla; l, labium. Maxilla composed of (1) cardo; (2) stipes bearing (3) 
undifferentiated galea and lacinia, and (4) palpus. 

Fig. 7. The frontal sclerite of head of larva with antennas and mandibles as removed from fig. 7. lr., 
labrum. 

Figs. 8 and 9 represent dorsal and lateral views of the posterior abdominal segment of the larva, (a) 
anus ; (b) dorsal sclerite; tr., lateral tracheal trunks; v., their vestibules (these are seen through the dorsal 
sclerite and open below its posterior ridge) ; (c) the lateral appendages or so-called " cerci." 

Plate II. 

Fig. 10. Dorsal view of penis. The abdominal segments are only partially extended. In numbering the 
abdominal segments I have followed Peytoureau. 7' and 8', the tergites of these segments ; 7, 8, and 9, 
the sternites ; a., accessory lobe of penis. ( x 60.) 

Fig. 11. Ventral view of penis, the connective tissue regions between segments 7 and 8 and 8 and 9 
being fully extended. Lettering as in fig. 10. ( x 60.) 

Fig. 12. Ventral view of 9 genital armature. 7 and 8, the abdominal segments, (a) Base of one 
of the spinnerets (b) ; [for b' see text] ; (c) the ventral grooved plate ; (x) one of the two pivots upon 
which the spinnerets swing ; (y) the chitinised lateral portion of segment 8. ( x 60.) 

Fig. 13. Lateral view of the $ genital armature, both vagina and rectum being prolapsed. Lettering 
as in fig. 12. ( x 60.) 

Fig. 14. Ventral view of ? genital armature in a state of rest showing the spinnerets retracted. 

Plate III. 

Figs. 15-31. The development of the egg. 

Fig. 15, 2-3 days stage, showing primitive streak (p.s.) and commencement of the amniotic cavity 
(am.c.) at posterior end of egg. j;er., peri vitelline space. 

Figs. 16 and 17, 3-4 days stage, a./., amniotic folds. The amnion (am.) has grown over the jtosterior 
part of the primitive streak towards the head, the lateral folds of which (hi.) are to be seen backed by the 
lateral folds of the amnion, bl., blastopore. 

Fig. 18. 5 days stage. (There is a stage wanting between this and that of the last figure owing to 
the three earlier drawings having been taken from an egg at a different time of year showing a different rate 
of development.) The embryo is seen bent backwards round the yolk, the appendages being visible as buds. 

Figs. 19 and 20. 6 and 7 days stages showing the gradual shortening of the embryo within the egg 
and the development of the appendages. 

Fig. 21. 8 days stage. The embryo showing the ruptured amnion twisted on the back to form the 
dorsal organ (d.o.). 



340 



THE LIFE-HISTORY OF HYDROBIUS FUSOIPES, L. 



Fig. 22. Ventral view of same embryo. 

Figs. 23 and 24. Lateral and ventral views of the 9th day stage showing the ventral curvature of the 
enil>i 

Fig. 25. 10 days stage showing the segmentation of the appendages and the commencement of the 
formation of the lateral tracheal trunks (tr.). 

Fig. 26. 11 days stage. The first sign of the heart (Id.) and of the muscular system. 

Fig. 27. 12 days stage. The pigmentation of the lateral tracheal trunks. 

Fig. 28. 13 days stage. The building up of the tracheal system. 

Fig. 29. The 14th day, some hours before the embryo is to emerge. Note the amount of free space 
within the shell. 

Figs. 30 and 31. The same egg shortly before and within a few minutes of the escape of the embryo. 
p.o., pulsating organ. The tracheal papillae ("setose gills") of the second thoracic and first abdominal 
segments are seen in fig. 30, and all these are visible in fig. 31. Note the tightly packed embryo in fig. 31. 

Figs. 32 and 33. The right and left mandibles of the imago seen from above. 

Fig. 34. The right antenna of the <$ of the imago. 




ins. Roy. Soc. Edin r - Vol. XLVII. 

F. Balfour Browne : Life History of Hydrobius fuscipes, L. — Plate I . 






fTTarlaue AErskine, Lith.Edin 1 " 



1 qs. Roy. Soc. Edin* Vol. XL VII. 

F. Balfour Browne: Life History of Hydrobius fuscipes, L. — Plate II. 







M'Farlane fcErskme Licb Edin r 



,. 



ns. Roy. Soc. Edin r - y ol X LVII. 

F. Balfour Browne: Life History of Hydrobius fuscipes, L.— Plate III. 






22. 





24. 





29. 




p.o. 



At. 




tt. 





d.o. 





kt. 





Z&v 



32. 



*V 



V 
33. 



/fY\f ~\ 




KFarMiu * Inline. Libh.Edir 



( 341 ) 



^AGE 


PAGE 

C. Effects on the Cerebro-Spinal Nervous System 364 




D. Effects on Skeletal Muscle .... 374 


341 


E. Effects on the Circulation .... 376 




F. Effects on Respiration 407 


349 


G. Summary 408 


351 


Explanation of Plates 409 



XV. — Strophanthus sarmentosus : its Pharmacological Action and its Use as an 
Arrow-poison. By Sir Thomas R. Fraser, M.D., F.R.SS. L. & E., Professor 
of Materia Medica in the University of Edinburgh, and Alister T. Mackenzie, 
M.A., M.B., Ch.B., Carnegie Research Scholar. (With Eleven Plates.) 

(MS. received July 2, 1909. Read May 3, 1909. Issued separately February 8, 1910.) 

CONTENTS. 

Introductory. — Sources of Materials used, Follicles 

and Seeds, Extract, Arrow - poison, 

Poisoned Arrows 

Pharmacological Action — 

A. Lethality or Toxic Power .... 

B. General Effects 

In the course of an endeavour, which was successful only after a number of years, 
to obtain specimens for the purpose of identifying the species of Strophanthus that 
produces the smooth seeds which had been chemically and pharmacologically investigated 
by one of us several years ago, the follicles, flowers and other parts of a number 
of different species of Strophanthus were obtained from Africa. Among them, 
S. sarmentosus was represented, and in the course of time a sufficient quantity of seeds 
of this plant was collected to allow an examination to be undertaken of their chemistry 
and pharmacology. 

In this communication, a description will be given chiefly of the pharmacological 
portion of the investigation, with only such brief reference to botanical and chemical 
facts as may be useful for the identification of the seeds, and of the preparations from 
them that were used in the experiments. Further, as there has come into our posses- 
sion much information regarding the use of the seeds of this plant as an arrow-poison, 
and an unusually complete collection of illustrative specimens, a few statements will 
also be made on this subject. It has a special interest inasmuch that arrows 
poisoned with these seeds are used against British troops in the punitive expeditions 
that are still found to be necessary in Nigeria and other parts of West Africa. 

The first specimens of S. sarmentosus that were obtained consisted of a few 
follicles from Gambia, in West Africa, sent to one of us in 1893, by Sir Robert 
Llewelyn, the then Governor of that Colony. From Gambia, also, specimens of 
flowers as well as of follicles and seeds were received in 1903, collected by Dr Dutton, 
of the Liverpool School of Tropical Medicine. In the same year Sir William 
Macgregor, K.C.M.G., sent some seeds and flowers from Lagos, of which Colony he 
was at that time Governor. In 1905, a large number of follicles was received from 
the Acting-Secretary to the Northern Nigeria Administration, and they were followed, 

TRANS. ROY. SOC. EDIN, VOL. XLVII. PART II. (NO. 15). 52 



342 SIR THOMAS R. FRASER AND MR ALISTER T. MACKENZIE ON 

in 1906, by several follicles and a small quantity of seeds collected by Dr John M. 
Dalziel, Colonial Medical Officer of Zungeru, Northern Nigeria. As these were 
supplemented, in 1908, by a large supply of follicles and seeds, also collected by 
Dr Dalziel. ample material had now been obtained for the purposes of investigation. 

We have to express our obligations to the above colonial officials, and especially 
to Dr Dalziel, as well as to the Colonial Office in London, by whom the co-operation 
of several of the above gentlemen and of the Administration of Northern Nigeria 
was secured. 

In order to obtain official confirmation of botanical identification, the whole, or 
specimens of each, of the above consignments were submitted to the authorities at 
the Royal Botanical Gardens, Kew, and their confirmation, and especially that of 
Dr Stapf, author of The Flora of Tropical Africa, may be accepted as a guarantee 
that the seeds used in this investigation were those of S. sarmentosus. 

In addition to the above-mentioned localities, this species of Strophanthus has also 
been found in Senegal,* Senegambia and Guinea t in West Africa, and a variety of 
the species, distinguished as S. verrucosus, in Zanzibar and districts adjoining 
Mombassa and Delagoa Bay in Eastern Africa.]; The plant has also received the 
names of S. Senegambia, S. pendulus and S. laurifolius. In the Lagos territory, the 
native name of " Isa- (or Isha-) kekere" has been given to it, that of " Kwankwanni" 
or " Kunkunni " in the Zungeru district of Northern Nigeria, and that of " Kunna 
Nombo " (shuttle vine) in the Gambia Colony. 

Follicles. — The follicles, which are generally fusiform in shape and have a shallow 
bifid apex, were received in a dry state, and either entirely closed or opened in 
various degrees at their ventral surfaces. In a considerable number (a) the carpel 
coverings are retained in their natural state ; and in others (b) the epi- and meso- 
carpels have been scraped off, no doubt to facilitate the drying of the contained seeds. 

(a) The non-scraped follicles are rough on the surface, with fine, longitudinal, linear 
projections ; and are of an irregularly mixed brown and green colour, the lower parts 
being chiefly brown and the upper parts chiefly green. In three or four out of 120 
examined, there was a number of small, slightly elevated, pale brown spots, about \ to 
4 mm. in length. The follicles vary in dimensions from 180 to 312 mm. (about 7 to 
12^- inches) in length, and from 20 mm. to 50 mm. (about ^ to 2 inches) in greatest 
width, the greater number being of or about the larger of these dimensions. The 
weight varies from 24 to 96 grams (about f to 3f ounces), the usual weight being 
about 60 or 70 grams (about 2^ to 2^- ounces), although a considerable number are 
heavier than the latter weights. 

(b) The scraped follicles are of a very pale fawn colour, and many of them have 
darker longitudinal markings, where the mesocarp has not entirely been removed. 

; * Payrau, Eecherches sur les Strophanthus, 1900, p. 86. 
t Stapf, Flora of Tropical Africa, vol. iv., 1904, p. 180. 
I Payrau, loc. cit., p. 86. 



STROPHANTHUS SARMENTOSUS : ITS PHARMACOLOGICAL ACTION. 343 

Their dimensions range between 180 mm. and 320 mm. in length (7-^ and 12^ inches), 
and 25 and 28 mm. in greatest width (l and 1-| inch) ; and their weight between 
11 and 39 grams (f to If ounces), the greater number approximating to the larger 
dimensions and the higher weight. 

Seeds. — The seeds, which are provided with the plumose appendages common to 
the genus, are narrow-ovate or spindle-shaped. They vary in size from 10x3 to 
13x4 mm. (about -4 x "12 to '5 x "15 inch), the greatest number being about 
12x4 mm. (-A7x'l5 inch); and they vary in weight from 0*014 to 0'028 gram 
(*22 to - 44 grain), having an average weight of about 0"024 gram ("37 or f of a 
grain). They are of a dark fawn colour and have a velvety appearance caused by the 
short brown hairs with which they are covered,* and their appearance is distinctive 
when contrasted with that of the seeds of many species of Strophanthus with which 
we have had an opportunity of comparing them. The seeds have a bitter taste, and, 
when reduced to powder, they have a feeble woody odour with slight acridity. 

When the surface of a transverse section is moistened with 80 per cent, sulphuric 
acid, a faint red tint appears round the cotyledons, in a few minutes the cotyledons 
assume a dull purple and the albumin a red tint, and, ultimately, the cotyledons 
become blue, and remain so for several hours. t When 80 per cent, sulphuric acid is 
applied to the seed divested of its testa, a lavender tinge is produced, which passes 
into violet and then into pale blue. Both of these reactions markedly contrast with 
the green colour produced in the seed of S. hispidus. Hydrochloric acid applied to 
a transverse section produces a greenish-blue colour in the interior of the seed, 
whereas the interior of the seed of S. hispidus is unaffected by this reagent. 

Extract of Seeds. — For the purpose of determining the pharmacological action of 
S. sarmentosus we used an alcoholic extract of the mature seeds. It was prepared 
by reducing the seeds to a fine powder, percolating the powder, in the first place, 
with ether so long as ether removed anything, and then percolating the residual 
powder with 90 per cent, alcohol in the proportion of five parts of alcohol to one part 
of the original powdered seeds. The etherial and alcohol percolates were separately 
distilled to a small bulk and evaporated at a low temperature on a water-bath, and the 
latter percolate finally dried in vacuo over sulphuric acid. By this process there was 
obtained 34 "7 per cent, of ether extract and 23 - 5 per cent, of alcohol extract. 

The latter extract, with which chiefly our experiments were made, is a brittle, 
shining, yellowish-brown and slightly hygroscopic substance, freely soluble in water 
and moderately so in 90 per cent, alcohol. The watery solution of the extract is 
strongly bitter in taste and slightly acid in reaction. It contains an active principle 

* A few of the seeds that had protruded to the outside of opened-up follicles, and many of the seeds that had 
been removed from the follicles and dried in Africa, are of a very pale fawn or deep cream colour, having become 
blanched, apparently, by exposure to weather. 

+ On submitting seeds to Mr Holmes, of the Pharmaceutical Society of Great Britain, who originated this test and 
has applied it to the seeds of many species of Strophanthus, he states that he obtained very similar colour-changes, 
and he expresses the opinion "that there can be no doubt, therefore, that your seeds are those of Strophanthus 
sarmentosus." 



344 SIR THOMAS R. FRASER AND MR ALISTER T. MACKENZIE ON 

having the chemical properties of a glucoside. In many of its colour reactions it 
closely resembles the extract of S. hispidus, as these have been described by one 
of us in a paper communicated to this Society.* It differs, however, in several 
colour reactions. For example, in the case of the dry extract, strong sulphuric acid 
produces a dark brown which slowly becomes violet; and dilute nitric acid a light 
yellow, which remains unchanged when the temperature is raised to 130° F. In the 
case of a 2 per cent, solution in water, phosphomolybdic acid causes an abundant 
yellow precipitate, which is dissolved by heating, and the solution on cooling becomes 
dark blue in colour with a greyish deposit ; and the solution of the extract is made 
bright yellow by the addition of molybdate of ammonium. 

Arrow-poisons. — In common with the seeds of several other species of Strophanthus, 
those of S. sarmentosus are employed in producing an arrow-poison, which is in 
extensive use in the districts of Africa where the plant occurs. Dr Dalziel has placed 
at our disposal specimens of this arrow-poison, the preparation of which he has had an 
opportunity of witnessing at Kantagora and at Zungeru, in Northern Nigeria. Accord- 
ing to Dr Dalziel, the natives place about a quarter of a pound of coarsely ground 
S. sarmentosus t seeds in an iron pot with about two pints of water and, after 
heating to the boiling point, add a small quantity of a mixed powder containing 
five ingredients, chiefly the bark and roots of several other plants. After boiling for 
a few minutes, the contents of the pot are percolated through a rough filter of twigs, 
leaves and husks, and the pea-soup-like percolate is again boiled until it has acquired 
the consistence of a thin syrup. At Kantagora, the powdered head of a small venom- 
ous serpent and a few drops of cock's blood are also added to the poison. The arrow- 
heads are poisoned by being several times dipped into this syrupy liquid, and at 
Zungeru they are finally dipped into some " animal substance." It is stated that 
a poisoned arrow-head may be used several times without renewal of the poison. The 
poison is a brownish-grey, grumous fluid of thin syrupy consistence, and holding 
minute, flaky, solid particles in suspension. After being at rest for several weeks, a 
small quantity of a brown, clear liquid collects at the upper part, and a considerable 
cohesive, grey sediment at the bottom. This sediment was found to be insoluble 
in water. 

When the thoroughly mixed poison was dried, it became a dark brown, brittle and 
slightly hygroscopic solid, 10 c.c. of the original poison yielding 10'14 grams of solid 
substance. When this dried poison was tested with the reagents giving fairly 
distinctive colour-changes with the extract of S. sarmentosus seeds, no conclusive 
results were obtained, no doubt because of the many ingredients, other than 
Strophanthus extract, which are present in the arrow-poison. 

The poison in its original form, the liquid that spontaneously separates from it, and 
the several times washed sediment were each tested in regard to lethality and effects 

* Trantactions of the Royal Society of Edinburgh, vol. xxxv., part iv., 1890, pp. 1001-3. 

t Some of these seeds were secured by Dr Dalziel, and on examination were found to be seeds of S. sarmentosus. 



STROPHANTHUS SARMENTOSUS : ITS PHARMACOLOGICAL ACTION. 345 

on the heart. While the former two were found to be active, though less so than the 
extract from the seeds, the third (sediment) appeared to be inert. The experiments are 
epitomised at pp. 347 and 348. 

There has also been sent by Sir Frederick Lugard, K.C.M.G., High Commissioner 
of Northern Nigeria, a description, obtained from a " professional native poison-maker " 
by Dr Gr. W. Thompstone, P.M.O., of the manner in which the poison is prepared. 
This coincides very closely with the description furnished by Dr Dalziel, even to the 
extent of there being used in the preparation, besides Strophanthus seeds, parts of five 
plants, the blood of a " red fowl " and, occasionally, the head of a venomous serpent. 
The " animal substance " referred to in Dr Dalziel's description of the manufacture of 
the arrow-poison is, according to this account, obtained, though only rarely, by 
sticking the arrows into the abdomen of a dead man and leaving them there for three 
days. The native name of the arrow-poison is " G-unguma," and arrows poisoned with 
it are in use at Zaria, Kantagora, Bida, Zungeru and Lapai. While, however, at each 
place, Strophanthus seeds are known to enter into its composition, it was not known if 
S. sarmentosus is the species, or the only species. The poison appears to be less fluid 
than that whose preparation was witnessed by Dr Dalziel, and, further, it is applied 
by being smeared on the arrow-heads with a feather and not by several times dipping 
the heads into the poison. It is said to remain active for three years, and to cause 
death, according to one account, in not less than twelve hours, and, in another account, 
in not less than thirty minutes " after a mere scratch." 

Poisoned Arrows from Sir Frederick Lugard (Plate I., figs. 4a and 46). — Arrows 
have not been received upon which had been applied the poison whose preparation had 
been observed by Dr Dalziel, but in 1904, Sir Frederick Lugard most kindly sent 
to one of us four leather quivers containing a large number of poisoned arrows obtained 
for him by Dr Leuer, one of the medical officers of Northern Nigeria. These arrows 
are not feathered. They vary in length from 685 to 787 mm. (27 to 31 inches), a 
large proportion of them being 762 mm. (30 inches). The shaft consists of bamboo 
cane, about 8 mm. ( T 5 g- of an inch) in diameter, which is notched at one extremity 
for receiving the bow-string, while about 38 to 51 mm. {\\ to 2 inches) of the iron 
stem of the arrow-head is inserted into the other extremity without being tied or fixed 
with any adhesive substance. In the majority of the arrows, the head is ovate in 
shape, and in others it is lanceolate (see fig. 4, Plate I.). The iron arrow-head with 
its stem is about 178 to 200 mm. (7 or 8 inches) in length, the stem being 128 to 165 
mm. (5 to 6^ inches) in length, and the head 25 to 50 mm. (1 to 2 inches) in length 
and 12 to 22 mm. (|- to f inch) in greatest width, a common width being about 19 mm. 
(f of an inch), and on each surface of the head one wing is concave and the other 
convex. From the base of each wing, a pointed barb projects downwards at an angle 
of about 20 degrees with the long axis of the arrow. The barbs are 11 to 15 mm. long 
(f to f inch), and 2 to 2 "5 mm. (y 1 ^ to -^ inch) thick at their attachment to the arrow- 
head. In two of the arrows that were received, there are also iron spikes projecting 



346 SIR THOMAS R. FRASER AND MR ALISTER T. MACKENZIE ON 

from each side of the iron stem, in one arrow two pairs and in the other three pairs 
(see fig. 4b, Plate I.). 

The long portion of the iron stem of the arrow-head that is not inserted into the 
cane shaft, and nearly always a considerable part also of the head itself, is thickly and 
unevenly covered with the poison, the iron stem having previously had a vegetable 
fibre wound spirally round it, apparently for the purpose of increasing the adhesion of 
the poison. 

When the arrows were first received the poison was slightly soft, but, in the 
course of time, it has become hard and brittle and greyish on the immediate 
surface, and almost black in the interior. The poison scraped from two arrow-heads, 
the one small and the other large, weighed 2*5 and 4*5 grams (38 "5 and 69 - 4 grains) 
respectively. 

On shaking the powdered poison in a test-tube with a little distilled water, a pale 
brown, slightly viscous fluid was produced, in which were numerous nearly white, 
flaky particles, which remained in suspension for several hours. Under similar treat- 
ment, 70 per cent, alcohol produced a brownish-yellow, clear solution, and 90 per cent, 
alcohol a very pale yellow solution. In each case, much the greater part of the poison 
remained undissolved. Neither ether nor chloroform appeared to exert any solvent 
action upon it. Although by chemical reagents definite proof could not be obtained 
that extract of S. sarmentosus seeds forms an ingredient of this poison — any more 
than with Dr Dalziel's arrow-poison, in which this extract is definitely known 
to be present — the manner of its preparation and the locality in which it is 
prepared are in favour of this being the case. This probability is strengthened by the 
results of experiments that we have made to determine its lethality and its effects on 
the heart, which are recorded at p. 348. 

Arrow-heads from the Chibbuh Hill Campaign. — We have also had an opportunity 
of examining two arrow-heads, obtained by Dr A. C. Parsons, M.O., East Bornu, and 
transmitted to one of us by the Acting High Commissioner of Northern Nigeria, which 
had been removed from the dead bodies of two members of the Chibbuh Hill Campaign 
of 1906-7. Both arrows are stated to have caused death in half an hour. They have 
the same characters as those sent by Sir Frederick Lugard. The lethality of the 
poison closely corresponds with that of the arrow-poison obtained by Dr Dalziel and 
of the poison on the arrows received from Sir Frederick Lugard, and it affects the 
heart in the same way (pp. 348 and 349). 

It may, therefore, be assumed as being probable that the seeds of S. sarmentosus 
form a main ingredient of the poison in these arrows also. 

With each of these arrow-poisons and with several of their products a number 
of experiments was made with the objects of determining their lethality and the 
nature of their action. Before being used, each substance was completely dried 
in vacuo over sulphuric acid. The experiments are summarised in the following 
Tables (I. to IV.). 



STROPHANTHUS SARMENTOSUS : ITS PHARMACOLOGICAL ACTION. 



347 



Table I. — Dr Dalziel's Arrow-Poison prepared from S. sarmentosus by the 
Natives of Northern Nigeria. 

Minimum-Lethal Dose for Rabbits by Subcutaneous Injection. 



Weight of 
Animal 
in grams. 


Actual Dose 
in grams. 


Dose per kilo- 
gram of Animal 
in grams. 


Result. 


2050 
1860 
1600 
1200 
1950 


0-00615 

0-00651 

00064 

0-009 

0039 


0-003 

0-0035 

0-004 

0-0075 

0-02 


Recovered. Slight effects. 
Reeovered. Distinct effects. 
Death in 5 hours 18 minutes. 
Death in 45 minutes. 
Death in 47 minutes. 



The minimum-lethal dose per kilogram of this poison by subcutaneous injection in 
rabbits was found to be # 004 gram, or 2f times larger than the minimum-lethal dose 
of the extract of S. sarmentosus. The arrow-poison was not completely soluble 
in water. The symptoms and post-mortem appearances caused by its administration 
were identical with those produced by the extract of S. sarmentosus, but they did not 
occur so soon after the injection. While death occurred within two and a half 
hours after the injection of the minimum-lethal dose of the extract, it did not occur 
until five hours after the injection of the minimum-lethal dose of this arrow-poison. 

Table II. — Liquid Part of Dr Dalziel's Arrow-Poison. 
Minimum- Lethal Dose for Rabbits by Subcuta?ieous Injection. 



Weight of 

Animal 
in grams. 


Actual Dose 
in grams. 


Dose per kilo- 
gram of Animal 
in grams. 


Result. 


2600 
1290 
1600 
1500 


0-0052 
00029 
0-004 
0-0045 


0-002 
0-00225 
0-0025 
0-003 


Recovery. Severe effects. 

Death in 2 hours 

Death in 3 hours. 

Death in 1 hour 10 minutes. 



The minimum- lethal dose per kilogram of this part of Dr Dalziel's arrow-poison by 
subcutaneous injection is, in rabbits, about 0-00225 gram per kilogram, or 1^ times the 
minimum-lethal dose of the extract of S. sarmentosus. Death occurred as rapidly 
after injection as in the case of the extract of S. sarmentosus, and similar 
symptoms and post-mortem appearances resulted. As it contains only soluble 
matter and presumably the Strophanthus products present in the arrow-poison, its 
greater toxicity is easily accounted for. The difference of lethality between it and the 
entire arrow-poison indicates that all or several of the other ingredients which enter 
into the composition of the entire poison diminish its lethality, and are introduced to 



348 



SIR THOMAS R. FRASER AND MR ALISTER T. MACKENZIE ON 



render the poison more viscous and adhesive, or under some misapprehension of their 
potency, or with a superstitious intention. 

That there are ingredients in the entire arrow-poison which diminish its lethality as 
contrasted with that of extract of S. sarmentosus was made apparent when the insoluble 
substance in the arrow-poison was separately examined. It was then found that when 
rabbits received by subcutaneous injection doses of this portion of Dr Dalziel's arrow- 
poison up to 0*02 gram per kilogram (or equivalent to 13 times the minimum-lethal 
dose of extract of S. sarmentosus), no definite symptoms were produced with even the 
maximum dose. 

Table III. — Northern Nigeria Poison from one of Sir Frederick Lugard's Arrows. 
Minimum- Lethal Dose for Rabbits by Subcutaneous Injection. 



Weight of 

Animal 

in grams. 


Actual Dose 
in grams. 


Dose per kilo- 
gram of Animal 
in grams. 


Result. 


1370 
1700 
1320 
1380 


0-00411 
0-00544 
0-00462 
0-00552 


0-003 
0-0032 
0-0035 
0-004 


Recovery. Slight effects. 
Recovery. Slight effects. 
Death in 3 hours. 
Death in 1 hour 17 minutes. 



The minimum-lethal dose of this poison for rabbits is the same as that of the poison 
sent by Dr Dalziel. The symptoms and post-mortem appearances are identical with 
those produced by extract of S. sarmentosus. 

When a dose of 0'004 gram per kilogram was injected subcutaneously into the 
flank of a rabbit, distinct symptoms were manifested in 50 minutes, and the animal 
died in 1 hour 17 minutes. When the same dose per kilogram (0'004 gram) was 
inserted in a dry state under the skin of the flank of a rabbit, distinct symptoms 
appeared in 50 minutes, and the animal died in 1 hour 5 minutes. As the poison 
removed from one of the smaller and one of the larger arrows weighed 2 5 and 4*5 
grams respectively, the one arrow carried sufficient poison to kill ten men and the 
other eighteen men. 



Table IV. — Arrow-Poison from one of the Chibbuh Hill Campaign Arrows. 
Minimum-Lethal Dose for Rabbits by Subcutaneous Injection. 



Weight of 

Animal 

in grams. 


Actual Dose 
in grams. 


Dose per kilo- 
gram of Animal 
in grams. 


Result. 


1700 
1300 
1650 


0051 

0-00455 

00066 


0003 

0-0035 

0-004 


Recovery. Slight effects. 
Recovery. Distinct effects. 
Death in 3 hours. 



STROPHANTHUS SARMENTOSUS : ITS PHARMACOLOGICAL ACTION. 



349 



The minimum-letha] dose of this arrow-poison is 2f times as great as that of extract 
of S. sarmentosus. The symptoms and post-mortem appearances are identical with 
those caused by extract of S. sarmentosus. 



In the investigation of the toxic power, the general effects and the effects on the 
more important functions and structures of the body of S. sarmentosus, we made use of an 
alcoholic extract of the seeds which had been deprived of all substances soluble in ether, 
and whose mode of preparation and chief characters have already been described (p. 343). 

A. Lethality or Toxic Power. 

The experiments made with the extract for the purpose of ascertaining the minimum 
quantity capable of producing death when administered by subcutaneous injection to 
frogs, rats, rabbits and cats, have been arranged in Tables V., VI., VII. and VIII. # 

Table V. — Minimum-Lethal Dose of Extract for Frogs (Rana iemporaria and R. esculenta). 



No. of 
Experiment. 


Weight of 
Animal in grams. 


Actual Dose 
in grams. 


Dose per kilo- 
gram of Animal 
in grams. 


Result. 


1 


30 R. esc. 


0000045 


0-0015 


Recovery. Slight effects. 


2 


24 R. temp. 


000006 


0-0025 


Recovery. Slight effects. 


3 


28 R. temp. 


0000077 


0-00275 


Recovery. Slight effects. 


4 


38 R. esc. 


0000114 


0-003 


Recovery. Distinct effects. 


5 


42 R. esc. 


0-000126 


0-003 


Recovery. Slight effects. 


6 


32 R. temp. 


0-0001024 


0-0032 


Recovery. Slight effects. 


7 


33 R. esc. 


0001089 


0-0033 


Recovery. Severe effects. 


8 


38 R. esc. 


0-0001254 


0-0033 


Death after 6 and before 21 hours. 


9 


41 R. esc. 


00001394 


00034 


Recovery. Severe effects. 


10 


41 R. temp. 


0-0001394 


0-0034 


Recovery. Distinct effects. 


11 


24 R. temp. 


0000084 


0-0035 


Death fifter 9 and before 10^ hours. 


12 


42 R. temp. 


0-000147 


0-0035 


Death in 3J hours. 


13 


42 R. esc. 


0-0001596 


0-0038 


Death in 4 hours. 


14 


30 R. temp. 


0-000112 


0-004 


Death in 2 hours 20 minutes. 


15 


34 R. esc. 


0-00017 


0-005 


Death in 1 hour 50 minutes. 



* In the case of warm-blooded animals — rats, rabbits and cats — the weight was taken and the extract was 
administered always at the end of the same interval of time (18 hours) subsequent to the last reception of food, in 
order to eliminate, as far as possible, variations in weight due to differences in the amount of the contents of the 
alimentary canal. 



TRANS. ROY. SOC. EDIN, TOL. XLVII. PART II. (NO. 15). 



53 



350 



SIR THOMAS R. FRASER AND MR ALISTER T. MACKENZIE ON 



Table VI. — Minimum-Lethal Dose of Extract for Rats. 



No. of 
Experiment. 


Weight of 
Animal in grams. 


Actual Dose 
in grams. 


Dose per kilo- 
gram of Animal 
in grams. 


Result. 


18 


240 


0-0072 


003 


Recovery. Very slight effects. 


19 


205 


0-041 


0-2 


Recovery. Slight effects. 


20 


185 


00555 


0-3 


Death in 4 days. 


21 


205 


0-123 


0-6 


Death in 4 days. 


22 


130 


0-0975 


0-75 


Death in 3 days. 


23 


180 


0-27 


1-5 


Death in 5 hours. 



Table VII. — Minimum-Lethal Dose of Extract for Rabbits. 



No. of 
Ex penmen t. 


Weight of 
Animal in grams. 


Actual Dose 
in grams. 


Dose per kilo- 
gram of Animal 
in grams. 


Result. 


24 


2000 


0-001 


0-0005 


Recovery. Very slight effects. 


25 


2000 


0-0015 


0-00075 


Recovery. Very slight effects. 


26 


2210 


0-00177 


0-0008 


Recovery. Very slight effects. 


27 


1940 


0001745 


00009 


Recovery. Very slight effects. 


28 


1870 


0-00187 


0-001 


Recovery. Distinct effects. 


29 


2500 


0-00275 


o-oon 


Recovery. Distinct effects. 


30 


1970 


0002364 


0-0012 


Recovery. Distinct effects. 


31 


1750 


0-002275 


0-0013 


Recovery. Distinct effects. 


32 


1640 


0-002296 


0-0014 


Recovery. Severe effects. 


33 


2330 


0-003262 


0-0014 


Recovery. Severe effects. 


34 


1460 


000219 


0-0015 


Death in 1 hour 25 minutes. 


35 


2200 


00033 


00015 


Death in 2 hours 2 minutes. 


36 


2150 


0-0043 


0002 


Death in 1 hour 17 minutes. 



Table VIII. — Minimum-Lethal Dose of Extract for Cats. 



No. of 
Experiment. 


Weight of 
Animal in grams. 


Actual Dose 
in grams. 


Dose per kilo- 
gram of Animal 
in grams. 


Result. 


38 
39 


2950 
2230 


0-00295 
0-00446 


o-ooi 

0-002 


Distinct effects. Recovery. 
Death in 1 hour 15 minutes. 



The following table (IX.) gives the results following the injection of the same extract 
into the marginal vein of the left ear in rabbits. 



STROPHANTHTJS SARMENTOSUS : ITS PHARMACOLOGICAL ACTION. 



351 



Table IX. — Intravenous Minimum-Lethal Dose of Extract for Rabbits. 



No. of 
Experiment. 


Weight of 
Animal in grams. 


Actual Dose 
in grams. 


Dose per kilo- 
gram of Animal 
in grams. 


Result. 


40 
41 
42 
43 
44 


2180 
2290 
1830 
1800 
2950 


0-00218 

0-00229 

0-002 

0-00216 

0-0059 


o-ooi 
o-ooi 

0-0011 
0-0012 
0-002 


Recovery. Severe effects. 
Recovery. Severe effects. 
Recovery. Severe effects. 
Death in 1 hour. 
Death in 4 minutes. 



From the foregoing Tables, it appears that, when the extract is injected into the 
subcutaneous tissues, the minimum-lethal dose per kilogram is in frogs about 0'0035 
gram; in rats, about 0"3 gram; in rabbits, about 0'0015 gram; and in cats, about 
0-002 gram.* 

In rabbits, the intravenous minimum-lethal dose is about 0"0012 gram per kilogram. 

With reference to the subcutaneous administration of the extract, in general terms 
rabbits and cats are equally susceptible, frogs are twice as resistant as, and rats 200 
times more resistant than, rabbits. 

A detailed account of several of the experiments in the Tables is given below to 
illustrate the symptoms produced by large non-lethal and by lethal doses, t 



B. General Effects. 

(a) Experiments on Cold-blooded Animals. — Frogs. 

Experiment VII. — Into the dorsal lymph-sac of a male frog [Rana esculenta) 
weighing 33 grams, 0"0001089 gram of extract, dissolved in Ringer's solution, was 
injected (equivalent to 0*0033 gram per kilogram, or ^ of minimum-lethal dose). Before 
the injection the throat and flank respirations were regular and at the rate of 20 per 10 
seconds; the nose reflex, conjunctival reflex and sacral reflex were acute. During 45 
minutes succeeding the injection, no symptoms appeared. In 1 hour 35 minutes, the 
throat respirations were 25 per 10 seconds, and irregular in amplitude; the flank 
respirations were 16 per 10 seconds and irregular in time; the flanks were distended ; 
the frog jumped well and would now remain on its back ; and the cardiac impacts were 
6 per 10 seconds. In 3 hours 45 minutes, the throat and flank respirations were 23 per 
10 seconds, regular and shallow ; the nose, conjunctival and sacral reflexes were acute ; 
the frog's head and throat rested on the floor of the tray ; its limbs were flexed ; it 
could jump well, but landed clumsily, and remained when placed on its back ; the flanks 

* The extract of S. sarmentosus used in these experiments is thus shown to possess only one-fifth of the lethal 
power in frogs and one-half in rabbits of the similarly prepared extract of S. hispidus which was used in the experi- 
ments with the latter substance made by one of us (Trans. E.S.E., vols. xxxv. and xxxvi., 1890 and 1891). 

t In this and in the other sections of the investigation, several other experiments were made, which gave results 
concording with those in the experiments that have been recorded. 



352 SIK THOMAS K. FRASER AND MR A.LISTER T. MACKENZIE ON 

were distended and the pupils large ; and the cardiac impacts were 6 per 10 seconds, 
regular and shallow. In 5 hours 27 minutes, the throat respirations were 23 per 10 
seconds ; the flank respirations were not visible unless the frog was disturbed ; the 
three reflexes previously referred to were acute ; the muzzle rested on the floor of tray ; 
the limbs were loosely flexed ; the frog could not jump well, the left anterior extremity 
appeared to be powerless ; the pupils were widely dilated ; and the animal remained on 
its back without any effort to recover the prone position, but when the abdomen was 
stroked, the prone position was rapidly resumed. 

Nineteen hours after the injection, the throat respirations were 22 per 10 seconds; there 
were no visible flank respirations while the frog was undisturbed ; the nose reflex was 
dull, the conjunctival reflex almost gone, but the sacral reflex was acute ; when the 
point of a blunt needle was lightly drawn across the ocular conjunctiva the frog breathed 
more deeply and moved feebly, but there was almost no movement of the eyelids ; the 
pupils were dilated and the limbs flexed ; when the posterior extremities were passively 
extended, there was feeble twitching of the toes, but the limbs were not drawn up, even 
on pressing the toes ; when the frog was laid on its back, the only movements resulting 
were slight tremors of the toes, and, when the abdomen was stroked, contraction of the 
muscles of the abdominal wall occurred ; the cardiac impacts were very distinct, and 3 
per 10 seconds ; and the flanks were much distended. The frog now weighed 33 grams. 
Twenty-four hours after injection, the conjunctival reflex was absent, and the condition 
of the muscular system as last noted. 

Twenty-seven hours after the injection, the frog was lying flaccid; the conjunctival 
reflex was present but sluggish ; the nose reflex and the sacral reflex were acute ; if one 
of the posterior extremities was extended passively it was drawn up again, but only 
after several efforts ; and the voluntary movements of the hind-limbs were feeble, and 
accompanied by twitching of the toes. 

Two days after the injection, the throat and flank respirations were 20 per 10 
seconds and regular ; the conjunctival reflex was acute ; the abdomen and flanks were 
much distended ; and the frog could jump, but not well, and at each effort some urine 
was ejected. Afterwards, the flanks became less distended, and the animal recovered 
the prone position rapidly when laid on the back. It now weighed 32 grams. 

On the following day, the only symptoms were slight distension of the flanks and a 
failure to jump as high as before the injection. The frog weighed 31 grams. 

Four days after the injection, the frog weighed 33 grams, and seemed in normal 
health. 

The laboratory temperature was 57° F. 

Experiment XI. — 0*000084 gram of extract dissolved in Ringer's solution was injected 
into the dorsal lymph-sac of a male frog (Rana temporaria) weighing 24 grams ( = 0*0035 
gram per kilogram, or the minimum-lethal dose). Before the injection, the throat and 
flank respirations were 28 per 10 seconds and regular in time ; the nose, conjunctival and 
Sacral reflexes were acute ; and the posture was normal and the animal jumped well. 



STROPHANTHUS SARMENTOSUS : ITS PHARMACOLOGICAL ACTION. 353 

Thirty minutes after the injec